Coding concept for coding bio-physiological signals and / or multi-channel digital signals

Abstract

A decoder for decoding a biophysiological signal from a data stream is presented, wherein the decoder is configured to decode the biophysiological signal from the data stream using an audio decoding scheme. Further aspects are described and also apply to other types of multi-channel digital signals.

Description

Technical Field

[0001] Embodiments of the present invention relate to apparatus and methods for encoding and decoding biological physiological signals and / or multi-channel digital signals. Background Technology

[0002] With advancements in digitization and data usage, the need for data compression is increasing for transmission and wider accessibility. For example, in the medical field, there may be a need to facilitate the exchange of medical data among healthcare professionals, and medical data may be observed and evaluated remotely by personnel or programs (such as medical staff, computer programs, and neural networks monitoring the physiological signals of multiple individuals). In other technological fields, similar data compression and / or data transmission may be necessary for data such as seismic measurements, weather measurements, long-term testing, etc.

[0003] This is achieved through the subject matter of the independent claims of this application.

[0004] Other embodiments of the invention are defined by the subject matter of the dependent claims of this application. Summary of the Invention

[0005] According to a first aspect of the invention, a decoder is provided for decoding biophysiological signals from a data stream, wherein the decoder is configured to decode biophysiological signals from the data stream using an audio decoding scheme.

[0006] It has been recognized that biophysiological signals can exhibit periodic patterns (e.g., signals associated with cardiac and / or brain activity), where such periodic patterns (e.g., distinct clustering within the spectrum of the biophysiological signal) can be accessed through compression using audio decoding schemes typically designed to leverage the fact that audio signals generally exhibit periodic patterns. Therefore, using audio decoding schemes allows for improved coding efficiency by utilizing the characteristics of biophysiological signals, even though they do not constitute typical audio signals. Consequently, such biophysiological signals can be encoded very efficiently, allowing for the storage of larger amounts of data (e.g., long-term measurements) and easier transmission. For example, biophysiological signals can be transmitted to healthcare professionals for remote observation and / or review by one or more physicians. Furthermore, biophysiological signals can be more easily included in a patient's digital files.

[0007] This invention describes an apparatus for encoding (or decoding) arbitrary digital waveform data (e.g., medical waveform data) into a bitstream. For example, suppose... For having A sequence of digital waveform data with 1 channel, where each channel ( ) including at a predetermined frequency The sampled digital storage is a sequence of sample values. Let... For channel The first in Each number stores the sample value, where the index... For example, with time Related. The stored sample value can be an integer or a floating-point value.

[0008] Several existing codecs for encoding and decoding digital waveform signals may be useful for encoding and decoding different types of digital waveform data. For example, an audio codec like Advanced Audio Coding (AAC) may be suitable for encoding and decoding neurophysiological signals from electroencephalography (EEG), electrocardiography (ECG), or electromyography (EMG). In other words, digital waveforms can be biophysiological waveforms.

[0009] It has been recognized that similar advantages can be applied to other technological fields that may also exhibit periodic patterns, such as seismic measurements, long-term stress testing (e.g., test objects exposed to vibration), weather measurements (e.g., periodicity of wind speed, light intensity, or water level), and stock market charts. Therefore, the further aspects described below can be applied to encoding biophysiological signals or other digital multichannel signals.

[0010] According to a second aspect of the invention, a decoder is provided for decoding a multi-channel digital signal from a data stream, wherein the decoder is configured to decode channel grouping information from the data stream and divide K' coded channels representing the multi-channel digital signal into Q sets according to the channel grouping information, each set having n i There are n encoding channels. i Indicates the number of encoded channels in set i, Q ≥ i ≥ 1, and uses a multi-channel decoding scheme to decode at least one set j from the data stream of Q sets, where n j ≥2.

[0011] It has been recognized that channel grouping can provide grouping that can improve encoding. For example, grouping can be provided between signals that have or do not have (or do not have sufficiently) periodic or waveform-like patterns. Furthermore, grouping can be provided between signals that have similar characteristics (e.g., similar in one or more aspects of amplitude, frequency, and offset). Therefore, different groups of signals can be encoded differently (e.g., aperiodic data can be encoded without an audio coding scheme and / or without entropy coding) and / or the different signals within the group can be used to improve encoding, for example, using inter-channel prediction and / or transforms with improved energy compression.

[0012] According to a third aspect of the invention, a decoder is provided for decoding a multi-channel digital signal from a data stream, wherein the decoder is configured to decode inter-channel delay information from the data stream, decode K channels of the multi-channel digital signal from the data stream, and delay the K channels to each other according to the inter-channel delay information.

[0013] It has been recognized that multiple channels of a digital signal may possess similar characteristics, which can be beneficial for coding techniques such as inter-channel prediction and transformation. Delaying K channels from each other can increase channel similarity, thereby improving coding efficiency. This offset in inter-channel delay can occur due to the use of multiple sensors positioned at different locations (e.g., on a patient) and / or the use of different measurement sensors. Particularly outside the realm of audio coding, compression-related techniques, including the use of inter-channel delay, may not be given much consideration.

[0014] According to a fourth aspect of the invention, a decoder for decoding a multi-channel digital signal from a data stream can be provided, wherein the decoder is configured to decode K' coded channels representing the multi-channel digital signal from the data stream and subject the K' coded channels to channel re-transformation, the channel re-transformation re-transforming the co-aligned (e.g., time-co-aligned) sample positions of the K' coded channels to obtain the K channels of the multi-channel digital signal.

[0015] It has been recognized that multiple channels may possess certain similarities, which (e.g., due to similar periodic patterns and / or amplitudes) can be utilized in transforms because the transform can be performed in domains that relate to and benefit from these similarities. For example, a transform using a waveform function such as the discrete cosine transform may result in clustering on certain spectra due to the periodicity of the signal. Such clustering can be used to represent data in a more compact form (e.g., energy compression).

[0016] According to a fifth aspect of the invention, a decoder is provided for decoding a multi-channel digital signal from a data stream, wherein the decoder is configured to decode channel permutation information from the data stream, decode N channels of the multi-channel digital signal from the data stream, and permutate K channels of the multi-channel digital signal according to the channel permutation information, thereby obtaining a multi-channel digital signal.

[0017] Permutations allow channels to be ordered according to patterns that facilitate grouping and sorting based on priority and / or purpose. Furthermore, permutations can allow for optimization of subsequent coding steps. For example, a permutation can allow channels to be correctly ordered to apply latency (e.g., ordered from lowest to highest latency and / or to reduce signaling mapping between latency and channels). Additionally, permutations can be determined to improve the coding efficiency of subsequent transforms (e.g., ordered according to spectral distribution). Permutations can also complement or compensate for subsequent grouping. For example, permutations can be performed for optimized latency and / or transforms, where grouping subsequently improves coding efficiency.

[0018] According to various aspects of the present invention, an encoder having features corresponding to the decoder described above is provided. Furthermore, methods executed or executable by such a decoder and encoder are also provided.

[0019] The problem to be solved can be defined as how to improve coding efficiency. Attached Figure Description

[0020] The accompanying drawings are not necessarily drawn to scale; their main purpose is usually to illustrate the principles of the invention. In the following description, various embodiments of the invention are described with reference to the following drawings:

[0021] Figure 1a A schematic example of a decoder for decoding biological physiological signals from a data stream is shown;

[0022] Figure 1b A schematic example of an encoder for encoding biological physiological signals into a data stream is shown;

[0023] Figure 2 A schematic view of an audio decoding scheme for decoding biological physiological signals from a data stream is shown;

[0024] Figure 3 A schematic view is shown of an audio decoding scheme for encoding biological physiological signals into a data stream;

[0025] Figure 4 An encoder for encoding multi-channel digital signals into a data stream and a decoder for decoding multi-channel digital signals from a data stream are shown.

[0026] Figure 5a A schematic view of a decoder used to decode multi-channel digital signals from a data stream is shown.

[0027] Figure 5b A schematic view of an encoder used to encode multi-channel digital signals into a data stream is shown.

[0028] Figure 6a A schematic view of a decoder used to decode multi-channel digital signals from a data stream is shown.

[0029] Figure 6b A schematic view of an encoder used to encode multi-channel digital signals into a data stream is shown.

[0030] Figure 7a A schematic example of a decoder for decoding multi-channel digital signals from a data stream is shown;

[0031] Figure 7b A schematic example of an encoder for encoding multi-channel digital signals into a data stream is shown;

[0032] Figure 8a A schematic view of a decoder used to decode multi-channel digital signals from a data stream is shown.

[0033] Figure 8b A schematic view of an encoder used to encode multi-channel digital signals into a data stream is shown.

[0034] Figure 9a It shows Figure 8a A schematic view of the decoder, where channel retransformation involves a series of partial channel retransformations;

[0035] Figure 9b It shows Figure 8b A schematic view of an encoder, where channel transformation involves a series of partial channel transformations;

[0036] Figure 10a It shows Figure 8a A schematic view of the decoder, wherein the decoder is configured to decode inter-channel delay information from the data stream;

[0037] Figure 10b It shows Figure 8b A schematic view of an encoder, wherein a multi-channel digital signal has N channels, and the encoder is configured to encode inter-channel delay information into the data stream;

[0038] Figure 11a It shows Figure 1a The decoder, wherein the biophysiological signals are multi-channel digital signals, and decoding the multi-channel digital signals involves grouping K' coded channels; and

[0039] Figure 11b It shows Figure 1b The encoder, wherein the biophysiological signal is a multi-channel digital signal, and encoding the multi-channel digital signal involves grouping K' encoded channels. Detailed Implementation

[0040] In the following description, identical or equivalent elements or elements having the same or equivalent functions are indicated by the same or equivalent reference numerals, even if they appear in different figures.

[0041] In the following description, numerous details are set forth to more fully explain embodiments of the invention. However, it will be apparent to those skilled in the art that embodiments of the invention can be practiced without these specific details. In other instances, well-known structures and devices are shown in block diagram form rather than in detail to avoid obscuring embodiments of the invention. Furthermore, unless explicitly stated otherwise, features of the different embodiments described herein can be combined with each other.

[0042] Figure 1a A schematic example of a decoder 100 for decoding biophysiological signals 102 from data stream 104 is shown, which is configured to decode biophysiological signals 102 from data stream 104 using an audio decoding scheme 106.

[0043] Figure 1b A schematic example of an encoder 200 for encoding a biophysiological signal 102 into a data stream 104 is shown, which is configured to encode the biophysiological signal 102 into the data stream 104 using an audio coding scheme 206.

[0044] Note that the decoded biophysiological signal 102 may be the same as the encoded biophysiological signal 102 encoded by the encoder 200, for example, in the case of lossless encoding. However, the biophysiological signal 102 and the encoded biophysiological signal 102 may differ, for example, due to lossy compression or other modifications (e.g., changes in sampling rate or amplitude during decoding). Similarly, Figure 1a and 1b The data stream 104 may be the same (e.g., at least in terms of payload) or may be different (e.g., due to packet loss and / or including transmission-related signaling). Audio coding schemes 106 and 206 may be the same (e.g., where encoder 200 is capable of performing the encoding aspects of audio coding scheme 106 / 206 and decoder 100 is capable of performing the decoding aspects of audio coding scheme 106 / 206), or audio coding scheme 106 may include a decoding aspect that allows decoding of data stream 104 encoded by the encoding aspects of audio coding scheme 206 (e.g., such that audio coding schemes 106 and 206 do not necessarily have to be the same, as long as compatibility between decoder 100 and encoder 200 is provided).

[0045] The biophysiological signal 102 may be an electrocardiographic signal, such as a signal obtained by at least one of electrocardiography, electroencephalography, and electromyography, such as medical waveform data, such as biophysiological waveform data, such as biomedical waveform data. The biophysiological signal 102 may represent the electrical activity and / or movement of the heart (e.g., a person's heartbeat). Alternatively or additionally, the biophysiological signal 102 may represent the electrical activity of the brain. The biophysiological signal 102 may be or may indicate a voltage determined by electrodes (e.g., which may be placed or positioned on a person's skin). The biophysiological signal 102 may be represented by an amplitude (e.g., voltage) that varies over time. The biophysiological signal 102 may include multiple samples (e.g., each sample indicating an amplitude value), wherein the samples are arranged in a temporal order (which may be different from or the same as the encoding order), for example, sampled at a fixed or variable sampling rate (e.g., between 10 Hz and 20 kHz).

[0046] The biophysiological signal 102 may have a single channel or multiple channels. The biophysiological signal 102 may include a channel for each electrode configured to determine a voltage (e.g., ten channels for a voltage determined by ten electrodes).

[0047] Data stream 104 may be defined solely by biophysiological signal 102 or may include other data. For example, data stream 104 may include data (e.g., one or more syntax elements) indicating one or more of the following: data type (e.g., whether data stream 104 is associated with a biophysiological signal or an audio signal), biophysiological signal type (e.g., electrocardiogram or electroencephalogram), electrode type, and patient information (e.g., one or more of age, name, gender, and insurance).

[0048] Audio decoding schemes 106 and 206 can be configured to perform lossy or lossless compression on multiple samples (e.g., the amplitude of an audio signal and / or the biophysiological signal 102). As will be discussed further below, audio decoding scheme 106 may be capable of predictive coding and / or performing sample transformation. In encoding and decoding, encoder 200 and decoder 100 can be configured to treat samples of the biophysiological signal 102 (e.g., the amplitude of a time increment) as samples of the audio signal (e.g., the amplitude of a time increment). As will be described below, encoder 200 and decoder 100 can be configured to disable the use of audio-specific techniques (e.g., psychoacoustic optimization), such as techniques related to limitations of human hearing perception. Disabling can be triggered by one or more syntax elements in the data stream (e.g., flags or syntax elements indicating the encoding of non-audio data and / or the encoding of the biophysiological signal) and / or by predetermined settings (e.g., when an encoding program is loaded on a medical device or computer).

[0049] For the sake of brevity, the references in the following text will be... Figure 2This describes a combination of several optional features of decoder 100. However, please note that... Figure 2 Any feature described herein may be provided individually or in combination with other features. Figure 2 Any other feature described herein or the entire disclosure herein (e.g., with any of the first to fifth aspects described in the foregoing invention description, for example with...) Figure 1a , 1b (or any one of the diagrams from 3 to 11b) are provided in any combination. Furthermore, any features disclosed in the reference decoder 100 may be correspondingly set in the encoder 200.

[0050] Figure 2 A schematic view of an audio decoding scheme for decoding biophysiological signals 10 from data stream 12 is shown. Any decoder 100 disclosed herein can be configured to use at least a portion of the audio decoding scheme (and vice versa for any encoder 200).

[0051] Audio decoding schemes may involve (e.g., in core decoder 31, which is part of decoder 100) transform-based audio decoding, including deriving a scaling factor and transform coefficients from data stream 12, spectrally shaping the transform coefficients using the scaling factor to obtain a shaped spectrum, retransforming the shaped spectrum to obtain an audio frame signal, and subjecting the audio frame signal to an overlap-add process (e.g., where the retransformation is linear); and / or linear predictive coding (LPC) audio decoding, including deriving LPC coefficients and information about the residual signal from the data stream, and using the LPC coefficients to subject the residual signal to LPC synthesis. It has been recognized that predictive coding is particularly advantageous for decoding biophysiological signals, as such signals often exhibit repetitive behavior, which allows for predictors with good accuracy.

[0052] Deriving the scaling factor from data stream 12 can involve entropy decoding of the scaling factor from data stream 12, or decoding LPC coefficients from data stream 12 and converting the LPC coefficients into the scaling factor. Entropy encoding can involve context selection based on previously decoded samples (e.g., their bits, such as saliency flags and / or symbol flags).

[0053] The biophysiological signal can be a multi-channel digital signal 10 (e.g., having two, three, four or more channels), and decoding the multi-channel digital signal from the data stream 12 using an audio decoding scheme includes: decoding K' encoded channels 14 representing the multi-channel digital signal from the data stream 12 by decoding n>1 encoded channels (where n≤K') from the data stream 12 using an audio decoding scheme, wherein the audio decoding scheme is a multi-channel audio decoding scheme 29, involving: decoding (31) an m-channel downmixed signal 32 (where 0<m<n, e.g., a dual-channel downmixed signal 32) from the data stream 12, and upmixing the m-channel downmixed signal 33 using auxiliary information 34 contained in the data stream to obtain n encoded channels. For example, the data stream 12 may include n=10 encoded channels 14 of ten electrodes, wherein the ten channels 14 are downmixed into a dual-channel downmixed signal 32, wherein the auxiliary information 34 allows upmixing the dual-channel downmixed signal 33 to obtain ten encoded channels 14.

[0054] Biophysiological signals may contain redundancy, such as similar periodicity and amplitude. Downmixing such signals can reduce this redundancy, thereby improving coding efficiency.

[0055] Biophysiological signals can be multi-channel digital signals, and decoding multi-channel digital signals from data stream 12 using an audio decoding scheme includes: grouping (e.g., also referred to as degrouping or ungrouping, since grouping can be revoked or degrouped by the corresponding encoder) the K' encoded channels 14 representing the multi-channel digital signals into Q sets 25 (e.g., such as...). Figure 2 As shown in the example, Q=2), each set has n i Each encoding channel (e.g., as shown) Figure 2 As shown, the first set 25 has channels 1 to n1, and the second set 25 has channels 1 to n2), n i Indicates the number of encoded channels in set i, Q ≥ i ≥ 1 (e.g., based on channel grouping information 40 decoded from data stream 12), and uses an audio decoding scheme to decode at least one set j from Q sets 25 (e.g., ...) from data stream 12. Figure 2 In the example, j = 1 or 2), for the set n j ≥2, where the audio decoding scheme is a multi-channel audio decoding scheme, involving: decoding m-channel downmixed signal 32 from data stream 12 (where m < n) j ), and use the auxiliary information 34 contained in data stream 12 to upmix the m-channel downmixed signal to obtain the n of set j. jThere are 14 encoded channels. For example, data stream 12 may include a dual-channel downmixing signal 32, wherein auxiliary information 34 contained in data stream 12 allows four channels (n1=4) to be derived for a first set j=1 and six channels (n2=6) to be derived for a second set j=2. However, any other number of downmixing channels, upmixing channels, and sets 25 may be provided. Channels 14 may be assigned to sets 25 based on one or more of signal similarity (e.g., similarity in one or more aspects of amplitude, frequency, and time offset), coding dependency, signal type, and electrode type.

[0056] The K' coded channels 14 can be grouped (or degrouped) based on signal similarity (e.g., amplitude, frequency, spectrum, and phase, or one or more of these). Alternatively or additionally, the K' coded channels 14 can be grouped (or degrouped) based on one or more priority criteria (e.g., according to spectral distribution, amplitude, or electrode). Different groups can be encoded according to different priorities (e.g., signals with higher frequencies can be transmitted last or not at all, for example, in cases of low transmission bandwidth).

[0057] Grouping (or degrouping) the K' coded channels 14 allows for the aggregation of similar channels 14, which can improve coding efficiency during downmixing. Grouping (or degrouping) the K' coded channels 14 according to priority allows for the definition of groups with higher priorities for signal decoding or signal display, and these groups can then be prioritized for transmission and / or for selecting error correction codes.

[0058] Decoder 100 (or audio decoding scheme) can be configured to subject K' coded channels of a multichannel digital signal to channel retransformation 16, which retransforms the co-aligned (time-co-aligned) sample positions 18 of the K' coded channels 14 to obtain K (retransformed) channels 20 of the multichannel digital signal (e.g., K=K' or K is different from K') (e.g., where the channel retransformation is linear).

[0059] Biophysiological signals can be well approximated by linear combinations of other channels. For example, biophysiological signals may have periodic patterns (e.g., the electrical activity of the heart), which can be well approximated by linear combinations of wave functions, which can concentrate signal energy. For example, the sample positions 18 of K' encoded channels 14 may be related to the frequency distribution of the wave function, where the K channels 20 of the multichannel digital signal obtained by retransformation (e.g., using Fourier transforms such as Discrete Cosine Transform (DCT) or Karhunen-Loève Transform (KLT)) can be (or form) the basis of the biophysiological signal.

[0060] Decoder 100 can be configured to derive channel retransformation from transform information 22 in data stream 12. For example, transform information 22 may indicate (or define) one or more transform matrices (and / or their inverses). Transform information 22 may define (or allow the deriving) of Karhunen-Loève transform (e.g., its matrix parameters and / or dimensions).

[0061] Decoder 100 can be configured to update channel retransforms (e.g., at each random access point RAP) based on transform information 22 in the data stream (such that different retransforms are used before and after the update). Transform information 22 may include differential and / or absolute values ​​for updating channel retransforms.

[0062] Channel retransformation can involve partial channel retransformation (see [link]). Figure 2 The sequence of transformations 24 in the diagram. For example, a total of twenty (e.g., K'=20) encoded channels 14 can be provided, where the first channel of the K channels 20 can be obtained by retransformation based only on a portion of the twenty co-aligned sample positions 18 (e.g., 16 sample positions out of 20 sample positions). The second channel of the K channels 20 can be obtained by retransformation based only on another portion of the twenty co-aligned sample positions 18 (or the same portion as the first channel) (e.g., 16 sample positions out of 20 sample positions). The third channel of the K channels 20 can be obtained by retransformation based on all twenty co-aligned sample positions 18. The use of partial channel retransformation can be combined with the use of full channel retransformation (i.e., using samples from all K' channels 14).

[0063] Partial channel retransformation can have different dimensions in terms of the number of retransformed channels 20, such that different encoded channels 14 are affected by different subsets of (re)transformations in the sequence of partial channel retransformations (where each subset can be a proper subset of or equal to the whole set containing all (re)transformations), or by the number of different (re)transformations or different (re)transformation subsequences. The subsets can be different from or the same as set 25.

[0064] The multi-channel decoding scheme can be designed to handle a maximum of MAX encoded channels 14, and the decoder 100 can be configured to group the K' encoded channels into Q sets, each set having n i n encoding channels, such that for each 0 < i < Q+1, n i≤MAX. For example, the decoding scheme can be designed to handle a maximum of six (e.g., MAX=8) encoded channels 14, and the data stream 12 includes 30 channels 14 (e.g., K'=30), wherein the decoder 100 can be configured to divide the 30 channels 14 into four sets 25, for example, three sets 25 each with eight channels 14, and a fourth set 25 with six channels 14. The maximum number MAX of encoded channels 14 can be predefined. Alternatively, the maximum number MAX of encoded channels 14 can be transmitted (e.g., once at the beginning of the data stream and / or at regular time intervals).

[0065] Decoder 100 can be configured to decode channel grouping information 40 from the data stream (e.g., as a basis for grouping K' encoded channels) and perform grouping using the channel grouping information. Channel grouping information 40 can indicate the value of the number n of channels in the Q sets; for example, decoder 100 can be configured to group channels into sets according to the encoding order. Alternatively or additionally, channel grouping information 40 can indicate the allocation of one or more (or all) channels 14 to corresponding sets 25 (e.g., using set IDs and / or set ID offsets).

[0066] The channel grouping information 40 may include syntax elements indicating the number n of coded channels in the Q sets, and / or the sequential channel order among the K' coded channels, using this sequential channel order to group the K' coded channels into Q sets 25 such that the coded channels 14 of different sets in the Q sets are non-interleaved along this sequential channel order. Therefore, the channel grouping information 40 may include auxiliary information for explicit signaling and / or implicit information conceived from this sequential channel order of the K' coded channels.

[0067] The auxiliary information may include one or more of the following: inter-channel coherence (ICC) data, channel level difference (CLD) data, and channel prediction coefficient (CPC) data.

[0068] Decoder 100 can be configured to decode inter-channel delay information 23 from the data stream and delay channels 20 (e.g., apply delays to one or more channels 20) to each other based on the inter-channel delay information 23. The inter-channel delay information 23 can indicate (or define) delays in time units (e.g., seconds or milliseconds) and / or sample units. The inter-channel delay information 23 can indicate absolute delays, such as relative to absolute time or global time or relative to predefined samples (e.g., the first sample, the first sample in a set of samples, the first sample in a time block, or the first block of a time channel block). Alternatively or additionally, the inter-channel delay information 23 can indicate relative delays, such as relative to a predefined channel (e.g., the first channel), a set of channels, or all channels. For example, a relative delay can be indicated relative to the first channel among K channels 20 or relative to the first channel (ordered by time) among K channels that exceed a threshold (e.g., exceeding a predetermined threshold indicating that other channels 20 are delayed).

[0069] For example, in Figure 2 In this process, K channels 20 are obtained by re-transformation 16 (e.g., from frequency domain to time domain) of K' channels 14, where inter-channel delay information 23 allows delay to be applied between channels 20.

[0070] Biophysiological signals may comprise multiple signals, which can be better compressed if their amplitudes are better aligned (e.g., due to more similar predictions and / or a better compressible sample distribution). For example, a biophysiological signal may comprise multiple voltage signals of cardiac electrical activity, where the signals may have similar periodicity but are offset from each other in the temporal direction (e.g., due to the distance of the electrodes relative to the heart and / or the asynchronous distribution of the heart's electrical activity). For encoding, compression can be improved by adding a delay to better align channel 20 (e.g., aligning the maximum amplitude of channel 20 in time). As a result, the delayed channels 20 are more similar to each other, which can reduce the complexity of the sample values ​​when transformed (at the encoder) to channel 14, potentially leading to better compression. Furthermore, inter-channel predictions can be improved. After decoding, the delay can be reintroduced to produce a time delay equal to (or close to) the original determined biophysiological signal. Note that the application of a delay to channel 20 does not necessarily require the use of a transformation and can also be used when the biophysiological signal is encoded (and transmitted) without a transformation (e.g., keeping the signal in the time domain without changing it to the frequency domain).

[0071] Decoder 100 can be configured to perform permutations based on channel permutation information 52 notified by signals in data stream 12 (see [link]). Figure 2The multichannel digital signal 10 is obtained by referring to channel 20 of the multichannel digital signal 10 (reference numeral 50). Channel permutation information 52 can indicate the mapping between channel 20 before and after permutation. Channel permutation information 52 can also indicate the offset between channels 20. Alternatively or additionally, the decoder 100 can be configured to permutate channel 20 of the multichannel digital signal 10 according to channel grouping information 40. Channel permutation information 52 and channel grouping information 40 can be transmitted at the beginning of the data stream and / or in the same (e.g., periodic) pattern.

[0072] Decoder 100 can be configured to extract data from substream 28 of the data stream associated with the corresponding set 25 in units of time interval 26 (e.g., defining a predetermined time interval, such as 16 milliseconds, 32 milliseconds, 64 milliseconds, or 128 milliseconds, or a predetermined amount of data, such as depending on the packet payload). Figure 2 The two sub-streams 28 of the two sets 25 shown decode each of the Q sets 25, such that each sub-stream 28 is composed of sub-stream portion 30 (see...). Figure 2 The sequence of the first sub-stream portion (reference number 30) marked as "1" is formed, the sequence of sub-stream portions having consecutive time intervals encoded therein in the set 25 associated with the corresponding sub-stream 28, and the sub-stream portions 30 of sub-stream 12 are interleaved such that sub-stream portions 30 whose time intervals overlap are adjacent and consecutive in the data stream 12, and precede the time interval of the subsequent sub-stream portion 30.

[0073] Figure 3 A schematic view is shown of a biophysiological signal 10 encoded into a data stream 12 using an audio decoding scheme. Any encoder 200 disclosed herein can be configured to use at least a portion of the audio encoding scheme. The disclosure herein regarding any encoder applies to (or includes) a corresponding decoder 100 capable of decoding a data stream encoded by such encoder 200.

[0074] The encoder 200 is configured to encode the biophysiological signal 10 into the data stream 12 using an audio encoding scheme.

[0075] The audio coding scheme may involve (e.g., in a core encoder 41 that may be part of encoder 200): transform-based audio coding, which includes inserting a scaling factor and transform coefficients into data stream 12, wherein the scaling factor is used to spectrally shape the transform coefficients to obtain a shaped spectrum, which is then transformed to obtain an audio frame signal, and the audio frame signal is subjected to an overlap-addition process to produce a reconstruction; or linear predictive coding (LPC) audio coding, which includes inserting LPC coefficients and information about the residual signal into the data stream, such that the residual signal is subjected to LPC synthesis using the LPC coefficients to produce a reconstruction.

[0076] The audio coding scheme may involve inserting a scaling factor into data stream 12, and may involve encoding the scaling factor entropy into data stream 12, or encoding LPC coefficients into data stream 12 to convert LPC coefficients into a scaling factor.

[0077] The biophysiological signal can be a multi-channel digital signal 10, and encoding the multi-channel digital signal 10 into the data stream 12 using an audio coding scheme can include: encoding K' coding channels 14 representing the multi-channel digital signal 10 into the data stream 12 by using an audio coding scheme to encode n>1 coding channels 14 (where n≤K') into the data stream 12, wherein the audio coding scheme is a multi-channel audio coding scheme 45, involving: encoding an m-channel upmixing signal 46 (which may correspond to an m-channel downmixing signal 32) into the data stream 12, where 0<m<n, and obtaining n coding channels 14 from the data stream 12 by using auxiliary information 34 encoded into the data stream 12 to downmix the m-channel upmixing signal 46 47.

[0078] The biophysiological signal can be a multi-channel digital signal 10, and encoding the multi-channel digital signal 10 into the data stream 12 using an audio coding scheme may include: grouping the K' coded channels 14 representing the multi-channel digital signal (see...). Figure 3 (Referencing reference number 44) represents Q sets, each with n... i 14,n encoding channels i The number of encoded channels in set i is indicated, Q ≥ i ≥ 1 (e.g., where the grouping can be derived based on the channel grouping information 40 encoded into data stream 12), and at least one set j of the Q sets (for the set n) is used with an audio coding scheme. j ≥2) Encode into data stream 12, wherein the audio encoding scheme is a multi-channel audio encoding scheme 45, involving: encoding the m-channel downmixed signal 46 (which may correspond to the m-channel downmixed signal 32) into data stream 12 so as to use the auxiliary information 34 contained in data stream 12 to upmix the m-channel downmixed signal 46 to obtain the n of set j. j 14 encoding channels.

[0079] The encoder 200 can be configured to subject K channels 20 of the multichannel digital signal 10 to a channel transformation 48 (e.g., the inverse transformation of retransformation 16, where the combination of channel transformation 48 and retransformation 16 produces a unit function), which transforms the co-aligned (time-co-aligned) sample positions of the K channels 20 to obtain K' encoded channels 14 of the multichannel digital signal 10 (e.g., K=K' or K is different from K', where the channel transformation is linear).

[0080] The encoder 200 can be configured to include transformation information 22 in the data stream 12, from which channel retransformation (e.g., channel retransformation 16) corresponding to channel transformation 48 can be derived.

[0081] Encoder 200 can be configured to change channel transform 48 and accordingly signal in the data stream to notify of the update of transform information (e.g., at each RAP, so that different transforms may be used before and after the update).

[0082] Channel transformation 48 may involve partial channel transformation (e.g., see...) Figure 3 The sequence of reference number 51 in the text.

[0083] The partial channel transform 51 can have different dimensions in terms of the number of transform channels 14, such that different encoded channels 20 are affected by different (heavy) transform subsets in the sequence of partial channel transforms (where each subset can be a proper subset of or equal to the whole set containing all (heavy) transforms), or by the number of different (heavy) transforms or different (heavy) transform subsequences.

[0084] The multi-channel encoding scheme can be designed to handle a maximum of MAX encoded channels, and the encoder 200 can be configured to group the K' encoded channels 14 representing the multi-channel digital signal 10 into Q sets, each set having n i There are 14 encoding channels, such that for each 0 < i < Q+1, n i ≤MAX.

[0085] Encoder 200 can be configured to encode channel grouping information 40 into data stream 12 (e.g., allowing K' encoded channels 14 to be grouped into groups 44), the channel grouping information indicating groups 44.

[0086] Channel grouping information 40 may include syntax elements indicating the number n of encoded channels in the Q sets. i (e.g., four channels 14 in the first set, eight channels 14 in the second set), and / or the sequential channel order among the K' encoded channels 14, grouping the K' encoded channels 14 into Q sets using the sequential channel order (e.g., by assigning set IDs to the encoded channels 14 and / or providing a mapping between the encoded channels 14 and sets 25), such that the encoded channels 14 in the different sets of the Q sets 25 are not interleaved along the sequential channel order.

[0087] The auxiliary information 34 may include one or more of ICC (inter-channel coherence) data, CLD (channel level difference) data, and CPC (channel prediction coefficient) data.

[0088] The encoder 200 can be configured to encode inter-channel delay information 23 into data stream 12, and to encode K channels 20 into data stream 12 in a state where K channels 20 are mutually delayed according to inter-channel delay information 23 (see reference numeral 35).

[0089] The encoder 200 can be configured to include channel permutation information 52 signaled in the data stream 12 and permutate channel 20 of the multi-channel digital signal 10 in order to encode channel 20 of the multi-channel digital signal 10 in a permuted state.

[0090] Channel permutation information 52 can describe the permutation between channel 20 of the multi-channel digital signal 10 encoded into the data stream 12 and a predetermined representation of the multi-channel digital signal 10 (see reference numeral 49).

[0091] The above description is extended by the presentation of the following additional embodiments. However, prior to this, the description continues to introduce possible frameworks or codecs into which the above embodiments and those described below can be built. However, many details described in this framework are optional when combined with any of the embodiments described above or subsequently. More precisely, the framework refers to... Figure 4 Describe it. Figure 4 An encoder 200 for encoding a multi-channel digital signal 10 into a data stream 12 and a decoder 100 for decoding the multi-channel digital signal 10 from the data stream are shown. Figure 4 The description should be considered as a presentation of new embodiments of this application, and any embodiments described above or subsequently are not necessarily related to them. Figure 4 When the decoder 100 or encoder 200 is combined, by adopting Figure 4 All details / features described or omitted Figure 4 It arises from some details / functions described. Sometimes, Figure 4 These “optional” features are explicitly identified as being optional in combination with respect to the embodiments described previously and subsequently, but the embodiments just mentioned above and those described previously / subsequently are not necessarily related. Figure 4 The possible combinations described should not be limited to these explicitly identified ones. Figure 4 A variant of this, where certain features are omitted.

[0092] exist Figure 4 In the diagram, the multi-channel digital signal 10 is illustrated using an array, and the samples are shown as small squares 19 (which can correspond to...). Figure 2 and 3 The sample position 18). Each row / line corresponds to a channel of the multi-channel digital signal 10 (e.g., Figure 2 and 3Channel 20, for example, with or without one or more of transforms, permutations, and time alignments. Each channel of signal 10 may have a corresponding channel ID associated with it. Figure 4 These channels are displayed in order of their channel IDs along the vertical axis 55; therefore, the vertical axis 55 corresponds to the "source" channel axis 55. The horizontal axis 56 corresponds to time, so samples 19 forming a column or horizontal alignment belong to a common moment. This set / column of time-coordinated samples 19 is shown in... Figure 4 The data is shown at 57. In the example in Figure 5, the multi-channel digital signal 10 has 32 channels (along the vertical axis 55), with 80 samples 19 per channel (along the horizontal axis 56), for a total of 32 × 80 = 2560 samples 19 (in the original domain 27). However, any other number of channels 20 and samples 19 can be used.

[0093] Therefore, each channel 20 forms a digital time-varying signal or a time / amplitude or time-to-amplitude signal (e.g., in...). Figure 4 (Each channel 20 has 80 values ​​within the shown time period). The multi-channel digital signal 10 may have been obtained through at least one of electrocardiogram, electroencephalogram, electromyography, or seismic measurement. In other words, the multi-channel digital signal may be biophysiological waveform data, such as electroencephalogram (EEG) signals, electrocardiogram (ECG) or electromyography (EMG) signals, or seismic waveform data. However, each channel / signal may be another type of waveform signal data, such as scalar media data, such as audio signals, and signal 10 may be a multi-channel audio signal.

[0094] Figure 4 This indicates the option that signal 10 is not directly encoded (i.e., in the original domain 27) but rather encoded in a so-called "encoded domain" 58, which may differ from the original domain 27 by one or more of the following: 1) channel transformation (e.g. Figure 3 The channel transformation shown is 48), 2) channel permutation (e.g. Figure 3 The permutations shown in 49) and 3) are time-intercorrelation channels aligned (e.g. Figure 3 The delay 35 is shown. If channel transformation is applied, the set or column 57 of samples at each sample time step is transformed from domain 27 to domain 58. Therefore, in domain 58, the sample spacing and time axis are the same as in domain 27, but the meaning of the channels is different; that is, the "source" channels of domain 27 become the transformed channels in domain 58. Therefore, Figure 4The vertical axis of domain 58 is represented as 59. Note that channel transformation may keep the number of channels unchanged, so the number of channels in domain 27 and domain 58 is the same (e.g., K'=K, also resulting in 32×80=2560 samples 63), but it may also use different methods (e.g., K'≠K). Typically, channel transformation aims to reduce redundancy and attempt to compress the channel energy onto a smaller number of channels in domain 58. As mentioned earlier, channel transformation is optional. Therefore, in general, the channels in domain 58 are called "coded channels" (which can correspond to...). Figure 2 and 3 Channel 14 (therefore hereinafter referred to as reference numeral 14) is distinguished from the “original” or “source” channel 20 of the digital signal 10 in domain 27. Permutation is also optional and can be used in conjunction with channel transformation or alone. If used in conjunction with channel transformation, permutation can be performed before the channel transformation (e.g., as...). Figure 3 (As shown) and / or performed afterward to permutate / sort the source channels 20 before the transformation and permutate the encoded channels after the channel transformation. The channel transformation may be a Discrete Cosine Transform (DCT), Discrete Sine Transform (DST), Fast Fourier Transform (FFT), or any other transformation. Time alignment is also optional and can be viewed as a constant time alignment between the source channels 20 or the encoded channels 14.

[0095] Figure 4 A module that performs one or more of channel transformations, channel permutations, and time alignment is represented as block 60. Auxiliary information 61 can be used to signal information about one or more of the following: 1) the channel transformation used, 2) permutation information between the source channel and / or the encoded channel, and 3) the mutual time alignment / delay between the source channel or the encoded channel, where time alignment may be limited to full sample accuracy. Auxiliary information 61 may include one or more of the following: transformation information 22, inter-channel delay information 23, auxiliary information 34, channel grouping information 40, and channel permutation information 52 (e.g., ...). Figure 2 and 3 (As shown). The corresponding block 62 in decoder 100 performs the inverse steps, namely, performing one or more of the following: 1) channel retransformation (e.g. Figure 2 1) Channel retransformation 16), 2) Repermutation of source channel 20 and / or encoding channel 14 (e.g. Figure 2 In the permutation 50), and 3) time realignment of the source or encoding channel (e.g. Figure 3 (Delay 35 in the text). Note that if no channel transformation occurs, encoded channel 14 is effectively equal to source channel 20, except that they are (optionally) time-aligned and / or ordered differently due to permutations. Block 62 may be controlled by the aforementioned auxiliary information 61.

[0096] Therefore, the "actual encoding" is related to encoding channel 14 in encoding field 58. In encoding field 58, encoding channel 14 is... Figure 4 The diagram represents lines or rows of samples 63, each extending along the time axis (or horizontal axis) 56. Encoded channels are stacked along the encoding channel axis 32 (possibly ordered according to their associated encoding IDs), thus forming an array of samples 63 (possibly corresponding to...). Figure 2 and Figure 3 (Samples from channel 20 in the sample). To reiterate, although... Figure 4 This describes the case where the number of source channels equals the number of encoding channels (in... Figure 4 The example depicts 32 source channels 20 and encoding channels 14, but the number may vary. Furthermore, if channel transformation is used, although there is no longer an explicit association between the source and encoding channels, a temporal association still exists: for each temporally co-occurring sample 57, there exists a temporally corresponding co-occurring encoded channel sample 63, where set 42 in field 58 is a column, and with mutual time alignment applied, set 42 may be a set of horizontally offset samples. Figure 4 In this case, it is exemplarily assumed that no such time alignment is performed, so sets 42 and 57 are both pure columns in the time / channel representation. For example, sample 63 in set 42 may define (or indicate) the coefficients of a transformation of a function along the vertical axis 55 that approximates sample 19 of set 57. Therefore, the individual samples 63 of set 42 may not necessarily define (e.g., by one-to-one mapping) the individual samples 19 of set 57, but the function along the vertical axis 55 can be defined by combination, for example, by retransformation.

[0097] Actual encoding is performed in units of so-called time blocks 65. The term "time block" 65 is used to denote the time portion of a multi-channel signal in domain 58, i.e., a set of encoded channels, specifically the time portion of a particular encoded channel. That is, for each time block 65, each encoded channel has a time block (e.g., block 40), as shown in the figure (taking time block 65c as an example). These time blocks 140 are distributed along a direction parallel to the time axis 58 (i.e., the width direction of the time block), and these blocks are co-located. Encoding is performed sequentially along these blocks 140 (or time blocks 140 or time channel blocks 140) following an encoding / decoding order 67. This encoding traverses blocks 140 one time block 65 at a time and along the channel order (e.g., the order of the encoded channels along axis 59) traversing the co-located blocks of the encoded channels 140 in time. This encoding / decoding order is... Figure 4The diagram is shown in Figure 67. That is, when time block 140 is the block currently to be encoded / decoded, the previously decoded / encoded time blocks include all previous time blocks of all encoded channels (e.g., all blocks 140 of time blocks 65a, 65a b) and time blocks that are co-occurring in time with the encoded channels preceding encoded channel 92 of time block 140 (e.g., all blocks 140 in time block 65c preceding the current encoded block, such as all blocks 140 above the current encoded block 140). These previously encoded / decoded time blocks and their samples are... Figure 4 The image is illustrated by shading. Note in this respect, in... Figure 4 To reduce the complexity of the diagram, only one time block 140 is explicitly illustrated. Therefore, in this specification, reference numeral 140 is sometimes used to indicate the current encoding / decoding time block, or to representatively represent all time blocks (e.g., the horizontal line of sample 63 extending along the width of blocks 65a, 65b, 65c). Furthermore, as... Figure 4 As shown, the way the signal 58 is divided into time blocks 65 and 140 may make these blocks 65 and 140 non-overlapping (e.g., forming independent or non-overlapping sub-divisions of the signal 58).

[0098] The actual encoding, performed in units of time blocks 140, is predictive. That is, encoder 200 includes block predictor 68, which predicts samples 63 of the currently encoded time block 140, thereby generating a prediction signal 69. The prediction residual 71, formed by the difference between the actual sample values ​​of time block 140 and the predicted samples of prediction signal 69, created by subtractor 73, is encoded into data stream 12 by residual encoder 70. Residual encoding in residual encoder 70 may or may not involve utilizing quantized encoding errors. In any case, block predictor 68 uses a reconfigurable version of a previously encoded time block (e.g., previously encoded block 140) to obtain prediction signal 69. The reconfigurable version 72 can be obtained at encoder 200 by residual decoder 74 and adder 78, wherein residual decoder 74 is used to reverse potential coding loss (e.g., to quantize using dequantization), which is reflected in the residual signal 76 encoded into data stream 12, and adder 78 adds the predicted signal 69 to the reconfigurable residual signal 80 obtained by residual decoder 74.

[0099] Decoder 100 decodes the encoded channels from data stream 12 in a corresponding manner, i.e., in units of time blocks 65 or 140, using predictive decoding. For this purpose, decoder 100 includes a residual decoder 82, an adder 84, and a block predictor 86, which correspond to and are interconnected in the same manner as elements 74, 78, and 68 of encoder 200. Specifically, residual decoder 82 derives a reconstructable residual signal 80 for the currently decoded time block 140 from residual signal 76 in data stream 12, and this signal is then added at adder 84 to a prediction signal 69 derived by block predictor 86 for time block 140 based on a reconstructed version 72 of previously decoded time blocks. Therefore, the output of adder 84 produces a reconstructed version 72 of the currently decoded time block 140, and this output becomes part of the decoded sample pool of previously decoded time blocks as the time blocks 140 of the encoded channel are traversed in this manner along encoding / decoding order 67 to reconstruct the encoded channel in encoding domain 58.

[0100] To achieve a high degree of random access capability, some time blocks 65 can be encoded in a random access manner, meaning that the encoding of the encoded channels 14 within them is independent of previous time blocks 65. For example, suppose time blocks 65b and 65e are random access time blocks. Then, no time channel block 140 in time block 65b or any time channel block 140 in time block 65e depends on any previous time block 140. For instance, no time block in time block 65a forms the basis of the encoding dependency of any time channel block 140 in time block 65b, and no time channel block 140 in time blocks 65a through 65d forms the basis of the encoding dependency of any time channel block 140 in time block 65e. Therefore, in other words, encoding dependencies are restricted so that they do not extend across the boundaries of random access time blocks 65b and 65e into any previous time block 65. This restriction may also apply to intermediate time blocks 65c to 65d between random access time blocks 65b and 65e, such that the intermediate blocks do not depend on any time blocks preceding the preceding time block (here, block 65b) in random access time blocks 65b and 65e. Accordingly, the forward time boundaries of random access time blocks 65b and 65e are... Figure 4 The middle part is represented by a thick line.

[0101] Furthermore, the encoding of encoding channel 14 may also interrupt or limit inter-channel dependencies by encoding one or more encoding channels as random access encoding channels, thereby limiting the encoding dependencies of these random access encoding channels (or even these random access encoding channels and intermediate encoding channels in between) so that they cannot extend across such random access encoding channels to any encoding channel in the channel order along axis 32 that precedes that random access encoding channel. Figure 4The diagram illustrates two such random access coded channels 88a and 88b and their associated inter-channel dependency boundaries.

[0102] The block predictor 68 of encoder 200 and the block predictor 86 of decoder 100 operate synchronously, meaning they generate the same prediction signal 69 based on previously encoded / decoded samples of previously encoded / decoded time blocks 140. On the encoder side 200, predictions for a specific time block 140 may be accompanied by one or more prediction parameters, or determined by these parameters. These prediction parameters 90 can be determined on the encoder side based on rate-distortion optimization. These prediction parameters 90 can be encoded into data stream 12 and decoded from data stream 12, and used by the block predictor 86 to perform the same prediction (e.g., prediction 69).

[0103] Encoder 200 and decoder 100 may support more than one prediction mode. For example, encoder 200 and decoder 100 may support intra-frame prediction mode (also known as block copy mode), according to which the current encoding / decoding time block 140 is based on the same encoding channel (in Figure 4In the example, the prediction is based on reconstructable sample values ​​of the previous encoded / decoded time blocks of encoding channel 92, to which the current encoded / decoded time block belongs. Alternatively or additionally, encoder 200 and decoder 100 may support an inter-frame prediction mode (also known as a cross-channel prediction mode), in which the current encoded / decoded time block 140 is predicted based on reconstructable sample values ​​of the previous encoded / decoded time blocks of the encoding channels preceding the encoding channel 92 to which the current encoded / decoded time block 140 belongs in the encoding order 32. Alternatively or additionally, a hybrid prediction mode may exist, in which the prediction signal 69 is obtained by simultaneously using reconstructed / reconstructable sample values ​​of the previous encoded / decoded time blocks of encoding channel 92 itself and reconstructed / reconstructable sample values ​​of the encoding channels preceding encoding channel 92 along axis 32 in the channel order. In addition, there may be some time blocks 140 that are encoded without any prediction at encoder 200 and decoded without any prediction at decoder 100, such as the first time block in slice 94 obtained by random access boundary 96 and random access channel boundary 98 separating the time blocks from each other. This corresponds to prediction signal 69 being set to zero, and this can form an additional mode called bypass mode. Additionally or alternatively, other modes may also exist, such as modes that derive a DC predictor or linear function predictor for block 69 based on the immediately preceding sample of block 140. Therefore, prediction parameter 90 may include a prediction mode flag or prediction mode indicator for the currently encoded / decoded time block 140 to indicate the prediction mode to be used for this currently encoded / decoded time block 140, and optionally include one or more parameters for parameterizing the prediction mode to be used for this currently encoded / decoded time block 140.

[0104] The aforementioned coding dependencies should not cross any boundaries 96 and 98, which stems not only from the sample prediction capabilities of the block predictors 68 and 86, but may also optionally stem from other mechanisms, such as parameter prediction—that is, predicting parameters of a particular time block 140, such as the aforementioned prediction parameter 90, based on the coding parameters transmitted in data stream 12 for any previous time block 140, or context deduction of any coding parameters (such as prediction parameter 90) or any other auxiliary information (such as auxiliary information 76 and 61) for context adaptive entropy coding / decoding time block 140 based on any coding parameters transmitted in data stream 12 for any previous time block.

[0105] In summary, encoder 200 converts the multichannel signal 10 to coding domain 58 and then encodes the coding channel 14 into data stream 12 using the block prediction method described above; decoder 100 decodes the coding channel of coding domain 58 from data stream 12 using the corresponding block prediction method, and then obtains the multichannel signal 10 in its original form 27 based on the coding channel 14 in coding domain 58 via segment 62. As previously mentioned, channel transformation (e.g., transformation between domains 27 and 58) is optional, and if not used, each sample 63 in coding domain 58 actually corresponds to sample 19 in the original domain 27. Furthermore, if time alignment is not used, each sample 63 corresponds precisely to sample 19 at the same time in the original domain 27, or in other words, all temporally co-occurring samples 63 in coding domain 58 remain temporally co-occurring in the original domain 27.

[0106] As mentioned above, Figure 4 This represents only a possible "framework," within which the foregoing embodiments and the embodiments described subsequently can be incorporated. Figure 4 Many modifications have been made, some of which may be mentioned in the following descriptions for certain embodiments described thereafter, but these modifications should be considered to apply also to other embodiments described thereafter. Furthermore, as a final point, and not to be construed as... Figure 4 Describing an exhaustive list of further possible modifications, it should be noted that the length of time block 65 may be... Figure 4 The length may not be as shown, meaning it may not be as... Figure 4Instead of remaining constant as shown, the length of the residual signal 80 can vary. For example, encoder 200 can determine the length of block 65 and signal the block length of block 65 (and the corresponding time block 140 of the encoding channel) within data stream 12. Furthermore, although not previously described, residual encoder 70 and residual decoder 82 may use transform encoding / decoding to transmit the residual signal 76 in data stream 12. That is, the residual signal 80 can be transmitted in data stream 12 in the form of a transform domain through the transform coefficients in the residual signal 76. The transform domain can be DCT, DST, or FFT. The transform can be non-overlapping, i.e., it may only transform the residual signal 80, and its re-transformation may only cover the residual signal 76 within block 140, and / or it may be unwindowed, i.e., the residual signal may be transformed without any transform window used for time-shaping of the residual signal 80 before transformation. The transform domain refers to the transform used by the encoder to transform the prediction residual signal 80 to be encoded from the time domain to the transform domain, and the corresponding retransformation from the transform domain to the time domain used by the decoder to derive the prediction residual signal 80. Alternatively, the transform can be selected from a set of available transforms, including, for example, one or more of the following: 1) one or more DCTs, 2) one or more DSTs, and 3) the identity transform, according to which the prediction residual signal 80 is directly encoded into the time domain in the data stream 12. Some deblocking processing can be used to avoid block artifacts. Alternatively, if an overlapping transform is used, the retransformation of the immediately preceding and following time blocks in the same coding channel can be overlapped and added together to completely reconstruct the residual signal 76 of the current time block 140. In addition to such transform (residual) coded blocks, there may also be temporal blocks 140 that additionally or alternatively use—besides the block predictions (which may be called primary predictions) performed by block predictors 68 / 86—to encode secondary per-sample predictions of the residual samples in residual block 71. This is done, for example, by predicting the residual of the current sample using residual values ​​previously decoded in blocks 71 or 80 in the sample encoding order, and then correcting it using secondary predicted residual samples decoded from data stream 12. The secondary predicted residual samples of such blocks can be encoded integrally into data stream 12 in transform domain form or per-sample in temporal domain form.

[0107] 2. Encoding biological physiological signals

[0108] This invention (e.g., encoder 200 and decoder 100) can convert a signal D (e.g., a biophysiological signal 102, or a multi-channel digital signal 10) into one or more sub-signals (e.g., channels 20, 14, and / or group 25) suitable for efficient (lossless or lossy) encoding and decoding using a given target encoder (e.g., Advanced Audio Coding (AAC)). Note that it may be advantageous to modify encoder 200 and / or decoder 100 of the target encoder to reflect the desired distortion behavior. For example, if lossy compression of signal D is desired and mean square error is minimized, it may be advantageous to disable psychoacoustic optimization in AAC. For example, encoder 200 and decoder 100 may be configured to use an audio coding scheme without psychoacoustic optimization (e.g., not using equal loudness profiles, auditory masking, and one or more of filtering and / or attenuation frequencies according to the limits of human hearing perception), or with one or more (or all) psychoacoustic optimizations disabled.

[0109] Figure 1a and 2 a shows a decoder 100 that decodes biophysiological signals 102 (e.g., multi-channel digital signals 10) from data stream 104 and an encoder 200 that encodes biophysiological signals 102 (e.g., multi-channel digital signals 10) into data stream 104, wherein Figures 2 to 4 The example illustrates a decoder 100 and encoder 200 with additional (optional) features, such as channel-level segmentation with optional channel reordering (e.g. Figure 2 and 3 The groups shown are 21 and 44), and the inter-channel delay (e.g.) Figure 3 The delay 35 based on inter-channel delay information 23), and the linear combination of codes (e.g. Figure 2 and Figure 3 The transformations and retransformations shown (48, 16), and the multi-layer linear combination of codes (e.g.) Figure 2 and 3 Partial channel (re)transformation 24, 51). These other features are particularly advantageous for encoding biological physiological signals 102, for example, because such signals may exhibit periodic characteristics, delays exist between multiple signals (e.g. due to measurement or bodily function itself), and have similar signal patterns that can be used for energy compression.

[0110] However, these features can also be beneficial in themselves, not necessarily in the context of encoding biophysiological signals. Below, embodiments according to various aspects are described, implemented independently of the aspects encoding biophysiological signals and other aspects. Therefore, any aspect can be implemented alone (i.e., not in combination with other aspects) or in combination with one or more other aspects (e.g., all or only some). Any disclosure below relating to multichannel digital signal encoding can be applied to one or more of biophysiological signals, seismic measurements (e.g., seismic waveform data), stock market data, and weather data.

[0111] Sub-signals (e.g., channels 20, 14, and / or set 25) can be encoded using the target codec, and the resulting so-called sub-bitstreams (e.g.) Figure 2 and 3 The sub-stream portion 30 depicted in the diagram can be combined into a single bitstream (e.g., bitstream 12). For decoding such a bitstream, the sub-bitstream is decoded from the bitstream using the target codec, and D (or its distorted version) can be reconstructed.

[0112] The following sections introduce several different methods for converting D into one or more sub-signals and combining related sub-bitstreams into a single bitstream.

[0113] 2.1 Channel-by-channel splitting with optional channel reordering

[0114] In this invention, a signal D (e.g., a media sample, such as an audio sample, a biophysiological signal 102, or a seismic measurement) can be converted into one or more digital waveform sub-signals, such that each channel of D (e.g., channels 14, 20) is contained in (at least) one digital waveform sub-signal. This technique can be used to reduce the number of channels in each sub-signal to a level supported by the target codec (e.g., grouping eight channels 14 into a set). The sub-signals are then encoded using the target codec, and the resulting sub-bitstreams are stored in the bitstream (e.g., sub-stream 28 in bitstream 12, for example, in the form of sub-stream portion 30). For each sub-bitstream contained in the bitstream (e.g., sub-stream 28 in bitstream 12, for example, in the form of sub-stream portion 30), information about how the channels contained in the sub-bitstream map to the channels of D can optionally be encoded into the bitstream (e.g., using channel grouping information 40). Note that this technique allows the creation of sub-signals containing arbitrary subsets of the channels of D and in arbitrary order. In this way, channels that are similar (e.g., in one or more aspects of amplitude, frequency, spectrum, and offset) can be grouped together. Depending on the capabilities of the target codec, encoding similar channels within the same sub-signal (e.g., similar channels 14 or 20 grouped in the same set 25) may be beneficial for coding efficiency. This could be the case, for example, when the target codec employs prediction from one channel to the others, or joint stereo coding of the sum and difference of two channel signals (also known as center-side) combined with optional prediction coding of the difference signal using the (original or reconstructed) sum signal.

[0115] In a preferred embodiment, the K channels of D (e.g., K channels 20 or K' channels 14) are divided into Q sub-signals (e.g., ... Figure 2 and 3 As shown, for two sets with Q=2, the channel indices associated with the substream are consecutive and increasing (e.g., as shown in the figure). Figure 2 and 3 As shown, 1 to n1 and 1 to n2), and where Q is a predefined number (e.g., predefined as two, three, four or more). This allows for particularly efficient signaling of the mapping between the channels of D and the sub-signals. For each sub-signal, it may be sufficient to encode (or infer) information about the number of channels contained (e.g., composed or formed by channel grouping information 40) (e.g., n1 and n2) and information about the channel index of the first channel of D contained in the sub-signal.

[0116] Figure 5a A schematic diagram of a decoder 100 for decoding a multi-channel digital signal 10 from a data stream 12 is shown. The decoder is configured to decode channel packet information 40 from the data stream 12 and to decode K' encoded channels 14 (or as shown) representing the multi-channel digital signal 10. Figure 2Channel 20 shown is grouped into Q sets 25 according to channel grouping information 40 (e.g., Figure 2 and 5a (two sets in the set), each set has n i There are n encoding channels. i Indicates the number of encoded channels in set i, Q≥i≥1, and decodes at least one set j (n) from data stream 12 using multi-channel decoding scheme 81. j ≥2).

[0117] Figure 5a The decoder 100 shown may additionally include any features of any decoder 100 disclosed herein (e.g., alone or in any combination), for example, referring to other figures (e.g. Figure 2 and Figure 4 ).

[0118] Channel grouping information 40 may include an indication of the number n of coded channels in the Q sets 25. i The syntax elements, and / or the sequential channel order among K' encoded channels 14 (or K channels 20, e.g., without transformation), are used to group the K' encoded channels 14 into Q sets 25, such that the encoded channels 14 in different sets of the Q sets are not interleaved along the sequential channel order, and one or more encoded channels of a set are consecutive in the sequential channel order. For example, all channels 14 of the first set may be followed by all channels 14 of the second set, and the second set may optionally be followed by all channels of other sets 25.

[0119] Decoder 100 can be configured to decode each set j (n) of Q sets 25 from data stream 12 using a multi-channel decoding scheme 71 (which may include or constitute part of any multi-channel decoding scheme disclosed herein, such as multi-channel audio decoding scheme 29). j ≥2).

[0120] Figure 5a Other features of the decoder 100 will be referenced below. Figure 2 and Figure 4 The description notes that using an audio decoding scheme is optional. (The above is about...) Figure 2 and Figure 4 The description can be applied in part or in whole. Figure 5a The decoder 100 shown.

[0121] A multi-channel decoding scheme 71 (e.g., a multi-channel audio decoding scheme 29) may involve: decoding m-channel downmixed signals 32 (m < n) from data stream 12. j ), and by using the auxiliary information 34 contained in data stream 12, the m-channel downmixed signal 32 is upmixed 33 to obtain nj Each coded channel. Auxiliary information 34 may include one or more of the following: ICC (Inter-channel coherence) data, CLD (Channel level difference) data, and CPC (Channel prediction coefficient) data.

[0122] The multi-channel decoding scheme 71 can be designed to process a maximum of MAX encoded channels 14, and the decoder 100 can be configured to group the K' encoded channels 14 representing the multi-channel digital signal 10 into Q sets 25, each set having n i n encoding channels, such that for each 0 < i < Q+1, n i ≤MAX.

[0123] Decoder 100 can be configured to subject K' coded channels 14 representing a multi-channel digital signal 10 to a channel retransformation 16, which retransforms the co-aligned (time-co-aligned) sample positions 18 of the K' coded channels 14 to obtain K channels 20 of the multi-channel digital signal 10 (e.g., K = K' or K is different from K', e.g., where the channel retransformation is linear). Decoder 100 can be configured to derive the channel retransformation 16 from transform information 22 in data stream 12. Decoder 100 can be configured to update the channel retransformation 16 based on the transform information 22 in data stream 12 (e.g., at each RAP, e.g., such that different channel retransformations are used before and after the update). Channel retransformation 16 may involve partial channel retransformation (e.g., see...). Figure 2 The sequence of reference numeral 24 in the table. Partial channel retransformation 24 can have different dimensions in terms of the number of retransformed channels 20, such that different encoded channels 14 are affected by different (re)transform subsets (where each subset can be a proper subset of or equal to the whole set of all (re)transformations) or different numbers of (re)transformations or different (re)transform subsequences in the sequence of partial channel retransformation 24.

[0124] Multi-channel decoding scheme 71 may be or use a multi-channel audio decoding scheme (e.g., multi-channel audio decoding scheme 29). Multi-channel decoding scheme 29 may involve: transform-based decoding (e.g., see...) Figure 2Reference numeral 31 in the text) or Linear Predictive Coding (LPC) decoding; Transform-based decoding includes: deriving a scaling factor and transform coefficients from data stream 12, spectrally shaping the transform coefficients using the scaling factor to obtain a shaped spectrum, retransforming the shaped spectrum to obtain a frame signal, and subjecting the frame signal to an overlap-add process (e.g., where the retransform is linear); LPC decoding includes: deriving LPC coefficients and information about the residual signal from data stream 12, and performing LPC synthesis on the residual signal using the LPC coefficients. Deriving the scaling factor from data stream 12 may involve: entropy decoding the scaling factor from data stream 12, or decoding LPC coefficients from data stream 12 and converting the LPC coefficients into a scaling factor.

[0125] The multi-channel digital signal 71 may have K channels 20, and the decoder 100 may be configured to decode inter-channel delay information 23 from the data stream 12 and delay the K channels 20 to each other by 35 according to the inter-channel delay information 23. The multi-channel digital signal 10 may include (e.g., or) biophysiological signals (e.g., biophysiological signal 102). The decoder 100 may be configured to permutate the channels 20 of the multi-channel digital signal 10 according to the channel permutation information 52 signaled in the data stream 12 (see, for example, see...). Figure 2 Reference numeral 50 in the table is used to obtain the multi-channel digital signal 10. Channel permutation information 52 can describe the permutation of channel 20 in order to obtain a predetermined representation of the multi-channel digital signal 10.

[0126] Figure 5b A schematic diagram of an encoder 200 for encoding a multi-channel digital signal 10 into a data stream 12 is shown. The encoder is configured to encode channel grouping information into the data stream 12, using this channel grouping information to indicate that the K' encoded channels 14 of the multi-channel digital signal are grouped into Q sets, each set having n... i There are n encoding channels. i Indicates the number of encoded channels in set i, Q≥i≥1, and uses a multi-channel coding scheme 83 to encode at least one set j (n j ≥2) Encode into the data stream.

[0127] Note that encoder 200 can be configured to determine channel grouping information 40 by using rate-distortion (R / D) testing and aiming to extremize R / D-related metrics, or by determining one or more metrics for different settings of channel grouping information 40 and selecting settings that extremize the one or more metrics or a combination of metrics determined by the one or more metrics. Multichannel coding scheme 83 can be the same as multichannel coding scheme 81. Multichannel coding scheme 81 and multichannel coding scheme 83 can be the decoding and coding aspects of an overall multichannel coding scheme. Multichannel coding schemes 81 and 83 can be any multichannel coding scheme disclosed herein.

[0128] Encoder 200 can be configured to encode via a reference. Figure 5a The decoder 100 (or any other decoder 100 described herein) decodes the data stream 12. The encoder 200 may have the same characteristics as... Figure 5a The decoder 100 describes any corresponding function and / or feature.

[0129] According to another aspect, a decoder is provided, configured to decode channel permutation information, as described below.

[0130] Figure 6a A schematic diagram of a decoder 100 for decoding a multi-channel digital signal 10 from a data stream 12 is shown. The decoder is configured to: decode channel permutation information 52 from the data stream 10, decode N channels 20 of the multi-channel digital signal 10 from the data stream 12, and permutate K channels 20 of the multi-channel digital signal 10 according to the channel permutation information 52 to obtain the multi-channel digital signal 10.

[0131] Note that encoder 200 can be configured to determine channel permutation information 52 by using rate distortion (R / D) testing and aiming to extremize R / D related metrics, or by determining one or more metrics for different settings for channel permutation information and selecting settings that extremize the one or more metrics or the combination of metrics determined by these metrics.

[0132] Other features of decoder 100 will be referenced. Figure 2 and Figure 4 The example shown is described below.

[0133] Channel permutation information 52 can describe the permutations 50 of the N channels 20 to obtain a predetermined representation of the multi-channel digital signal 10. Decoder 100 can be configured to decode the N channels 20 of the multi-channel digital signal 10 from the data stream 12 using a multi-channel decoding scheme (e.g., any multi-channel decoding scheme 10 disclosed herein, such as biophysiological signal 102). Decoder 100 can be configured to group the K' encoded channels 14 representing the multi-channel digital signal 10 into Q sets 25, each set having n... i There are n encoding channels. i Indicates the number of encoded channels in set i, where Q ≥ i ≥ 1, and decodes at least one set j (n) from data stream 12 using multi-channel decoding scheme 81. j ≥2), to decode N channels 20 of the multi-channel digital signal 10 from data stream 12. Decoder 100 can be configured to decode each set j (n) of Q sets 25 from data stream 12 using multi-channel decoding scheme 81 (which can be any multi-channel decoding scheme disclosed herein). j ≥2). A multi-channel decoding scheme 81 may involve decoding m channels of downmixed signal 32 (m < n) from data stream 12. j ), and upmix the m-channel downmixed signal using auxiliary information 34 contained in data stream 12 to obtain n j Each coded channel 14. Auxiliary information 34 may include one or more of the following: ICC (Inter-channel coherence) data, CLD (Channel level difference) data, and CPC (Channel prediction coefficient) data.

[0134] The multi-channel decoding scheme 81 can be or uses a multi-channel audio decoding scheme (such as any multi-channel audio decoding scheme disclosed herein). The multi-channel decoding scheme 81 may involve transform-based decoding (see, for example, see...). Figure 2 Reference numeral 31) or Linear Predictive Coding (LPC) decoding, transform-based decoding includes: deriving a scaling factor and transform coefficients from the data stream, spectrally shaping the transform coefficients using the scaling factor to obtain a shaped spectrum, re-transforming the shaped spectrum to obtain a frame signal, and subjecting the frame signal to an overlap-add process (e.g., where the re-transform is linear). LPC decoding includes: deriving LPC coefficients and information about the residual signal from the data stream, and performing LPC synthesis on the residual signal using the LPC coefficients. Deriving the scaling factor from the data stream 12 may involve: decoding the scaling factor from the data stream entropy, or decoding the LPC coefficients from the data stream and converting the LPC coefficients into a scaling factor. The multi-channel decoding scheme 81 can be designed to handle a maximum number of MAX encoded channels to the greatest extent, and the decoder is configured to group the K' encoded channels representing the multi-channel digital signal 10 into Q sets, each set having n i n encoding channels, such that for each 0 < i < Q+1, ni ≤MAX.

[0135] Decoder 100 can be configured to decode from data stream 12 (see...) Figure 2 Reference numeral 29 in the diagram represents K' coded channels 14 of a multi-channel digital signal 10, and the K' coded channels 14 representing the multi-channel digital signal 10 undergo channel retransformation 16, which retransforms the temporally co-occurring sample positions 18 of the K' coded channels 14 to obtain K channels 20 of the multi-channel digital signal 10 (e.g., K=K' or K is different from K', e.g., where the channel retransformation is linear), to decode N channels 20 of the multi-channel digital signal 10 from the data stream 12. The decoder 100 can be configured to derive the channel retransformation 16 from the transform information 22 in the data stream 12. The decoder 100 can be configured to update the channel retransformation 16 based on the transform information 22 in the data stream 12 (e.g., at each RAP, e.g., such that different channel retransformations are used before and after the update). The channel retransformation 16 may involve partial channel retransformations (e.g., see...). Figure 2 The sequence of reference numeral 24 in the table. Partial channel retransformation 24 can have different dimensions in terms of the number of retransformed channels 20, such that different encoded channels 14 are affected by different subsets of (re)transformations in the partial channel retransformation 24 sequence (where each subset can be a proper subset of or equal to the whole set of all (re)transformations) or different numbers of (re)transformations or different (re)transformation subsequences.

[0136] Decoder 100 can be configured to decode inter-channel delay information 23 from data stream 12 and delay K channels 20 to each other by 35 according to inter-channel delay information 23. Multi-channel digital signal 10 may include (e.g., or) biophysiological signals (e.g., any biophysiological signals disclosed herein).

[0137] Figure 6b A schematic diagram of an encoder 200 for encoding a multi-channel digital signal 10 into a data stream 12 is shown. The encoder is configured to encode channel permutation information 52 into the data stream 12 and to encode N channels 20 of the multi-channel digital signal 10 into the data stream 12 in such a way that K channels 20 of the multi-channel digital signal 10 are permuted 50 according to the channel permutation information 52 to obtain the multi-channel digital signal 10.

[0138] The encoder 200 may have one or more additional features of any encoder 200 disclosed herein. Figure 6b The encoder 200 shown can be configured to encode by Figure 6a The decoder 100 shown decodes the data stream 12. The encoder 200 may include... Figure 6aThe features of the decoder 100 shown correspond to any encoder-side features.

[0139] 2.2 Inter-channel delay

[0140] Certain types of digital waveform signals can exhibit high correlations between channels. For example, spikes in an electrocardiogram (ECG) recording are caused by the operation of the human heart (e.g., voltage spikes indicating cardiac electrical activity). Therefore, they are likely to appear in different channels of the ECG recording in a similar manner at almost the same time points. However, depending on the distance of the ECG electrodes from the human heart, there may be (e.g., a constant) time delay between spikes in different channels. This can be a disadvantage for a target encoder that makes predictions between channels.

[0141] In this invention, sequence D (e.g. Figure 2 and 3 Channel 14 (or channel 20) shown can be converted into one or more digital waveform sub-signals, such that the sample sequence of each channel can be prefixed with a predefined number of (artificially created) padding sample values ​​(or any other form of delay, such as fixed or variable time duration), which correspond to the delay. Delay information associated with a particular channel can be transmitted in a bitstream (e.g., bitstream 12). The decoder can use this delay information (e.g., inter-channel delay information 23) to remove the padding sample values ​​(or any other form of delay, such as that included by the corresponding encoder 200) after decoding and recover (potentially distorted) the sequence D with the correct inter-channel delay.

[0142] Figure 7a A schematic example of a decoder 100 for decoding a multi-channel digital signal 10 (e.g., including or constituting a biophysiological signal) from a data stream 12 is shown. The decoder is configured to decode inter-channel delay information 23 from the data stream 12 and decode K channels 20 of the multi-channel digital signal 10 from the data stream 12 (e.g., using inter-channel dependencies, such as inter-channel prediction, M / S decoding), and delay the K channels 20 to each other according to the inter-channel delay information 23 (see reference numeral 35).

[0143] Figure 7a The decoder 100 shown may additionally include any features of any decoder 100 disclosed herein (e.g., alone or in any combination), for example, referring to other figures (e.g. Figure 2 and Figure 4 ). Figure 7a Other features of the decoder 100 will be referenced below. Figure 2 and Figure 4 The description notes that using an audio decoding scheme is optional. (The above is about...) Figure 2 and Figure 4The description can be applied in part or in whole. Figure 7a The decoder 100 shown.

[0144] Decoder 100 can be configured to decode K channels 20 using a multi-channel decoding scheme (e.g., a multi-channel audio decoding scheme and / or a spatial object audio decoding scheme). The multi-channel decoding scheme (e.g., an audio decoding scheme) may involve (see, for example, [reference needed]). Figure 2 Reference numeral 31) Transform-based multichannel (e.g., audio) decoding or linear predictive coding / decoding (LPC) decoding; transform-based multichannel decoding includes: obtaining a scaling factor and transform coefficients from data stream 12, spectrally shaping the transform coefficients using the scaling factor to obtain a shaped spectrum, retransforming the shaped spectrum to obtain a frame signal, and subjecting the frame signal to an overlap-addition process (e.g., where the retransformation is linear); LPC decoding includes: deriving LPC coefficients and information about the residual signal from the data stream, and performing LPC synthesis on the residual signal using the LPC coefficients. Deriving the scaling factor from data stream 12 may involve: entropy decoding the scaling factor from data stream 12, or decoding LPC coefficients from the data stream and converting the LPC coefficients into a scaling factor. Multichannel digital signals may include (e.g.,) biophysiological signals.

[0145] Decoding K channels 20 from data stream 12 may include using a multi-channel decoding scheme 29 to decode n>1 coded channels 14 representing a multi-channel digital signal 10 from data stream 12. This scheme involves: decoding (31) an m-channel downmixed signal 32 (0 < m < n) from data stream 12, and upmixing the m-channel downmixed signal 33 using auxiliary information 34 contained in data stream 12 to obtain n coded channels. Decoding K channels 20 from data stream 12 may include grouping the K' coded channels 14 representing the multi-channel digital signal 10 into Q sets 25, each set having n... i There are n encoding channels. i The number of encoded channels in set i is indicated, Q ≥ i ≥ 1 (e.g., based on channel grouping information 40 decoded from the data stream, such as K = K' or K is different from K'), and a multi-channel decoding scheme is used (e.g., an audio decoding scheme, such as involving decoding m-channel downmixed signal 32 (m < n) from data stream 12). j And by using the auxiliary information 34 contained in data stream 12 to upmix the m-channel downmixed signal, the n of set j is obtained. j The decoding scheme of 14 encoding channels decodes at least one set j (n) of Q sets 25 from data stream 12. j ≥2).

[0146] Decoding K channels 20 from data stream 12 may include using a multi-channel decoding scheme (e.g., an audio decoding scheme, such as involving: decoding m-channel downmixed signal 32 (m < n) from the data stream and upmixing the m-channel downmixed signal using auxiliary information 34 contained in the data stream to derive a decoding scheme for n encoded channels 14) to decode n ≤ K' encoded channels 14 representing multi-channel digital signals 10 from data stream 12.

[0147] Decoder 100 can be configured to perform channel retransformation 16 on K' coded channels 14 of a multi-channel digital signal 10, wherein the channel retransformation retransforms the temporally co-occurring sample positions 18 of the K' coded channels 14 to obtain K channels 20 of the multi-channel digital signal 10 (e.g., where channel retransformation 16 is linear). Decoder 100 can be configured to derive channel retransformation 16 from transform information 22 in data stream 12. Decoder 100 can be configured to update channel retransformation 16 based on transform information 22 in data stream 12 (e.g., at each RAP, such that different channel retransformations are used before and after the update). Channel retransformation may involve partial channel retransformation (see...). Figure 2 The sequence of reference numeral 24 in the table. Partial channel retransformation 24 can have different dimensions in terms of the number of retransformed channels 20, such that different encoded channels 14 are affected by different subsets of (re)transformation in the sequence of partial channel retransformation (where each subset can be a proper subset of or equal to the whole set of all (re)transformations) or different number of (re)transformations or different (re)transformation subsequences.

[0148] The multi-channel decoding scheme can be designed to handle a maximum of MAX encoded channels 14, and the decoder is configured to group 21 the K' encoded channels 14 representing the multi-channel digital signal 10 into Q sets 25, each set having n i n encoding channels, such that for each 0 < i < Q+1, n i ≤MAX.

[0149] Decoder 100 can be configured to decode channel grouping information 40 from data stream 12 (e.g., as a basis for grouping K' encoded channels 14) and perform grouping 21 using channel grouping information 40. Channel grouping information 21 may include an indication n of the number of encoded channels in the Q sets 25. i The syntax elements, and / or the sequential channel order among the K' encoded channels 14, using the sequential channel order, the K' encoded channels 14 are grouped into Q sets 25 such that the encoded channels 14 in different sets of the Q sets are not interleaved along the sequential channel order, and one or more encoded channels 14 in a set are adjacent and consecutive in the sequential channel order.

[0150] The auxiliary information may include one or more of the following: ICC (Inter-Channel Coherence) data, CLD (Channel Level Difference) data, and CPC (Channel Prediction Coefficient) data. The decoder 100 may be configured to permutate K channels 20 of the multi-channel digital signal 10 according to the channel permutation information 52 signaled in the data stream 12, to obtain the multi-channel digital signal 10. The channel permutation information 52 may describe the permutations of the N channels 20 obtained through channel re-transformation 48 to obtain a predetermined representation of the multi-channel digital signal 10.

[0151] Figure 7b A schematic example of an encoder 200 for encoding a multi-channel digital signal 10 into a data stream 12 is shown. The encoder is configured to encode inter-channel delay information 23 into the data stream 12 and to encode K channels 20 of the multi-channel digital signal 10 into the data stream 12 in a state where the inter-channel delay information 23 is mutually delayed by 35.

[0152] Figure 7b The encoder 200 shown may have one or more additional features of any encoder 200 disclosed herein. Figure 7b The encoder 200 shown can be configured to encode by Figure 7a The decoder 100 shown decodes the data stream 12. The encoder 200 may include... Figure 7a The features of the decoder 100 shown correspond to any encoder-side features.

[0153] It should be noted that the delay 35 may be determined by the encoder 200 to balance the different audio frame boundary start positions of the individual channels 20 (optionally and / or in cases where the frame lengths used to encode the individual encoded channels 20 are different), thereby aligning the individual encoded channels 20 temporally to each other for easier or more efficient encoding, for example, by using an optimization scheme that minimizes the mutual similarity metric between channels 20. It should also be noted that, according to the example, all K channels 20 may have been uniformly sampled, i.e., uniform in terms of sampling frequency and sampling delay / time, or the encoder may resample the channels to be encoded to achieve this state. However, it is also possible that grouping and / or resampling is effective only for channels within a set (e.g., set 25), and not necessarily for channels belonging to different sets. Furthermore, sampling rate deviations such as one channel's sampling rate being an integer multiple of another channel are also allowed. In other embodiments, the encoder 200 may keep the sampling pattern of the channels 20 unchanged, regardless of mutual sampling time / frequency differences.

[0154] 2.3 Linear Combination of Decoding Channels

[0155] Certain types of digital waveform signals may contain channels (e.g., channels 20 and / or 14) that can be well approximated as linear combinations of other channels (or functions). Furthermore, digital waveform signals can also be transformed by employing linear transformations (e.g., Discrete Cosine Transform (DCT) or Karhunen-Loève Transform (KLT)) to concentrate signal energy.

[0156] In this invention, a sequence D (e.g., a multi-channel digital signal 10) can be converted into one or more digital waveform sub-signals in a manner that allows the decoder to reconstruct the channels of D by creating a linear combination (e.g., a weighted sum) of channels 14 decoded from the sub-bitstream. This may require encoding information (e.g., transform information 22) for each channel of D about how to compute the weighted sum of the channels contained in the sub-bitstream. For example, a weight value can be encoded in bitstream 12 for each channel of D and for each channel contained in the sub-bitstream (or one or more sets of channels 14). Note that the number of channels contained in the sub-bitstream does not need to match the number of channels in D (e.g., K ≠ K'). In fact, it may be advantageous in terms of compression efficiency if the number of channels contained in the sub-bitstream is less than the number of channels in D (e.g., K > K'). Note that the channels may have different sampling frequencies. If so, the linear combination may require resampling some channels contained in the sub-bitstream to achieve a common sampling frequency, such as the frequency of the channel to be reconstructed.

[0157] In a preferred embodiment, a predefined set of weight values ​​is defined for calculating linear combinations, and whether one of these sets should be used is indicated in bitstream 12, for example, by using flags or indices. This requires fewer bits than explicitly signaling all weights in the bitstream. For example, the weights of the inverse discrete cosine transform (IDCT) can be indicated by information encoded in bitstream 12, for example, by flags or indices.

[0158] Figure 8a A schematic diagram of a decoder 100 for decoding a multichannel digital signal 10 (e.g., a biophysiological signal, such as any multichannel digital signal 10 disclosed herein) from a data stream 12 is shown. The decoder is configured to decode (see reference numeral 29) K' coded channels 14 representing the multichannel digital signal 10 from the data stream 12 and perform a channel retransformation 16 on the K' coded channels 14, which retransforms the co-aligned (e.g., temporally co-located) sample positions 18 of the K' coded channels 14 to obtain K channels 20 of the multichannel digital signal 10.

[0159] Figure 8a The decoder 100 shown may also include any features of any decoder 100 disclosed herein (e.g., alone or in any combination), for example, referring to other figures (e.g. Figure 2 and Figure 4 ). Figure 8a Other features of the decoder 100 will be referenced below. Figure 2 and Figure 4 The description should be provided, noting that the use of an audio decoding scheme is optional. The above refers to... Figure 2 and Figure 4 The description can be applied in part or in whole. Figure 8a The decoder 100 shown.

[0160] The number of decoded channels K' can be equal to the number of channels 20 obtained K (e.g., K = K'). The channel retransformation 16 can be linear. The decoder 100 can be configured to derive the channel retransformation 16 from the transform information 22 in the data stream 12.

[0161] Decoder 100 can be configured to update channel retransform 16 based on transform information 22 in the data stream (e.g., at each RAP, such as causing different channel retransforms to be used before and after the update). Decoder 100 can be configured to decode K' encoded channels 14 of the multi-channel digital signal 10 from the data stream 12 using an audio decoding scheme (e.g., a multi-channel or spatial object audio decoding scheme, such as any audio decoding scheme disclosed herein). The audio decoding scheme can be (or uses) a multi-channel audio decoding scheme.

[0162] Decoder 100 can be configured to decode K' encoded channels 14 of a multi-channel digital signal 10 from data stream 12 using a decoding scheme (e.g., Figure 2 Reference numeral 31 in the document refers to a decoding scheme involving transform-based decoding or linear predictive coding / decoding (LPC) decoding. Transform-based decoding includes: deriving a scaling factor and transform coefficients from data stream 12; spectrally shaping the transform coefficients using the scaling factor to obtain a shaped spectrum; re-transforming the shaped spectrum to obtain a frame signal; and subjecting the frame signal to an overlap-addition process (e.g., where the re-transformation could be inverse MDCT). LPC decoding includes: deriving LPC coefficients and information about the residual signal from data stream 12; and performing LPC synthesis on the residual signal using the LPC coefficients. Deriving the scaling factor from the data stream may include entropy decoding the scaling factor from data stream 12, or decoding the LPC coefficients from data stream 12 and converting the LPC coefficients into a scaling factor.

[0163] Decoder 100 can be configured to decode K' encoded channels 14 of a multi-channel digital signal 10 from data stream 12 using a multi-channel decoding scheme, the multi-channel decoding scheme including: decoding from data stream 12 (see...) Figure 2The m-channel downmixed signal 32 (where 0 < m < n) is obtained by upmixing the m-channel downmixed signal 32 using auxiliary information 34 contained in the data stream 12, resulting in n encoded channels 14 (where n ≤ K'). The decoder 100 can be configured to group the K' encoded channels 14 into Q sets 25, each set having n i There are n encoding channels, where n i Indicates the number of encoded channels in set i, Q ≥ i ≥ 1 (e.g., based on channel packet information 40 decoded from data stream 12), and uses a multi-channel decoding scheme to decode at least one set j from Q sets 25 of data stream 12 (e.g., each set j) (for the set n) j ≥2), the multi-channel decoding scheme involves: decoding m-channel downmixed signal 32 from data stream 12 (where m < n) j ), and by using auxiliary information 34 contained in data stream 12 to upmix the m-channel downmixed signal 32, the n of set j is obtained. j The 14 encoded channels (e.g., where m may differ among the sets 35 encoded using a multi-channel scheme; therefore, the symbol m) j (May be used).

[0164] The multi-channel decoding scheme can be designed to handle a maximum of MAX encoded channels, and the decoder 100 can be configured to group K' encoded channels 14 into Q sets 25, each set having n i n encoding channels, such that for each 0 < i < Q+1, n i ≤MAX. Auxiliary information 34 includes one or more of ICC (Inter-Channel Coherence) data, CLD (Channel Level Difference) data, and CPC (Channel Prediction Coefficient) data. Decoder 100 can be configured to decode channel grouping information 40 from data stream 12 (e.g., as a basis for grouping 21 of K' coded channels 14) and perform grouping 21 using channel grouping information 40. Channel grouping information 40 may include an indication n of the number of coded channels in the Q sets 25. i The syntax elements, and / or the sequential channel order among the K' encoded channels 14, according to the sequential channel order, the K' encoded channels 14 are grouped into Q sets 25 such that the encoded channels 14 of different sets in the Q sets 25 are not interleaved along the sequential channel order, and one or more encoded channels of the same set are adjacent to each other in the sequential channel order.

[0165] Decoder 100 can be configured to permutate K channels 20 of multichannel digital signal 10 according to channel permutation information 52 signaled in data stream 12 to obtain multichannel digital signal 10. Channel permutation information 52 can describe the permutations 50 of N channels 20 obtained through channel retransformation 16 to obtain a predetermined representation of multichannel digital signal 10. Multichannel digital signal 10 may include (e.g., or) biophysiological signal 102 (e.g., any biophysiological signal 102 disclosed herein).

[0166] Figure 8b A schematic diagram of an encoder 200 for encoding a multi-channel digital signal 10 into a data stream 12 is shown. The encoder is configured to perform a channel transformation 48 on K channels 20 of the multi-channel digital signal 10, which transforms the co-aligned (temporally co-located) sample positions 19 of the K channels 20 to obtain K' encoded channels 14, and encodes the K' encoded channels 14 representing the multi-channel digital signal 10 into the data stream 12.

[0167] The encoder 100 can be configured to resample one or more of the K channels 20 of the multi-channel digital signal 10 such that the K channels 20 or one or more sets of K channels 20 sample each other at the same frequency and the same sampling phase, and / or to sample each other at the same frequency that is an integer multiple of the common fundamental frequency and perform phase adjustment.

[0168] Figure 8b The encoder 200 shown may have one or more additional features of any encoder 200 disclosed herein. Figure 8b The encoder 200 shown can be configured to encode by Figure 8a The decoder 100 shown decodes the data stream 12. The encoder 200 may include... Figure 8a The features of the decoder 100 shown correspond to any encoder-side features.

[0169] 2.4 Multi-layer linear combination of decoding channels

[0170] If multiple linear transformations can be applied sequentially to a sequence (For example, a multi-channel digital signal 10), this can be advantageous. For instance, a DCT can be applied to the sequence, which typically compresses the signal energy into a small subset of the transformed channels (e.g., channel 14). Therefore, it might be a good idea to subsequently apply a KLT only to such a subset of the transformed channels with high signal energy. The resulting signal can, for example, consist of two sets of transformed channels (or any other number of channels). For example, one set of channels could be the result of applying a DCT followed by a KLT, and the other set of channels could be the result of applying only a DCT. Both sets (e.g., set 25) can then be encoded into a bitstream (e.g., bitstream 12). The decoder operates in reverse order. For example, it first applies an inverse KLT to the associated group of channels, and then applies an inverse DCT to all channels to reconstruct the signal. (Possibly distorted version).

[0171] In this invention, a bitstream (e.g., bitstream 12) may comprise sub-bitstreams generated through multiple linear transformations (e.g., partial transformation 24), and decoding of such bitstreams can be performed as follows. Let... In order to reconstruct (potentially distorted) sequences And the number of linear combination stages in continuous application. Let... For application of the first The sequence of digital waveform data preceding each linear combination stage. That is, It consists of channels decoded from the sub-bit stream (e.g., channel 14, before the first linear combination stage is applied). This is the result of applying the first linear combination stage, and so on. Ultimately, It is the result of the final linear combination stage, which corresponds to the (potentially distorted) reconstructed sequence. (For example, channel 20). Signal The number of channels for each They may be different.

[0172] In a preferred embodiment, the linear combination stage Input signal The channels are divided into two subsets. One subset is bypassed (e.g., by performing an identity transformation, such as not using the channels of said subset in stage i, but instead using them as input to stage i+1) to the linear combination stage. The output of one subset is transformed by performing a linear combination. Note that the linear combination stage can be interpreted as 2D matrix multiplication. The bypass channel corresponds to the row where only the diagonal elements are 1 and all other values ​​are 0. This technique reduces the number of weights required to compute the linear combination, which saves on weight signaling costs and computational complexity.

[0173] In another preferred embodiment, the weights associated with the linear combination stage are either explicitly signaled or selected from a predefined set of transforms (indicated in the bitstream of the selected transform).

[0174] In another preferred embodiment, the first linear combination stage is for Apply the first inverse transform (e.g., inverse KLT) to a subset (or all) of the channels, and bypass the remaining channels (if any) to The weights of the first inverse transform can be explicitly signaled, or selected from the set of first transforms and indicated in the bitstream. The second linear combination stage can... A subset (or all) of the channels are subjected to a second inverse transform (e.g., inverse DCT), and the remaining channels (if any) are bypassed. The weights of the second inverse transform can be explicitly signaled, or selected from the set of second transforms and indicated in the bitstream.

[0175] In another preferred embodiment, an offset value is associated with each weight value and is added to the result of multiplying the channel by the weight. The offset value can be explicitly signaled, for example, in the bitstream.

[0176] Figure 9a It shows Figure 8a A schematic diagram of decoder 100 is provided (but the principle can be applied to any decoder 100 disclosed herein), wherein channel retransformation 16 involves a sequence of partial channel retransformations 24-1 to 24-p (summarized by reference numeral 24). The partial channel retransformations 24 can have different dimensions in terms of the number of transformed channels, such that different encoded channels are affected by different subsets of (re)transformations in the sequence of partial channel retransformations 24 (where each subset can be a proper subset of or equal to the whole set containing all (re)transformations), the number of different (re)transformations, or different sequences of (re)transformations.

[0177] Figure 9b It shows Figure 8b A schematic diagram of encoder 200 (but the principle can be applied to any decoder 200 disclosed herein), wherein channel transformation 48 involves a sequence of partial channel transformations 51-p to 51-1 (summarized by reference numeral 51).

[0178] The partial channel transform 51 can have different dimensions in terms of the number of transform channels, so that different encoded channels are affected by different (heavy) transform subsets of the partial channel transform 51 (where each subset can be a proper subset of the whole set containing all (heavy) transforms or equal to the whole set), different (heavy) transform numbers, or different (heavy) transform subsequences.

[0179] 2.5 Combination of inter-channel delay and linear combination of decoding channels

[0180] For a signal to achieve good decorrelation through linear transformations along the channel direction, perfect channel alignment may be crucial (e.g., for coding efficiency). That is, if there are delays between channels, linear combination may not be able to decorrelate the signal.

[0181] In this invention, a time delay value (e.g., in terms of time and / or sampling, such as buffered samples) can be applied individually to each channel (e.g., all channels 20 or all channels 20 of a subset) to improve the decorrelation characteristics of the linear combination stage.

[0182] Figure 10a It shows Figure 8a A schematic diagram of decoder 100 (but this principle can be applied to any decoder 100 disclosed herein, such as...) Figure 9a The decoder 100 is configured to decode inter-channel delay information 23 from the data stream 12 (optionally, the K' encoded channels 14 of the multi-channel digital signal 10 can be decoded from the data stream 12 using inter-channel decoding dependencies, such as using inter-channel prediction), and to delay the K channels 20 to each other according to the inter-channel delay information 23.

[0183] Decoder 100 can be configured to delay each other 35 K channels 20 in a manner based on inter-channel delay information 23, such that samples from two channels in the K channels are subject to a mutual delay greater than one sampling interval (e.g., two, three, four or more sampling intervals).

[0184] Figure 10b It shows Figure 8b A schematic diagram of encoder 200 (but this principle can be applied to any decoder 200 disclosed herein, such as...). Figure 9b The encoder 100 is configured to encode inter-channel delay information 23 into the data stream 12 (optionally, the K' encoded channels 14 of the multi-channel digital signal 10 can be encoded into the data stream 12 using inter-channel decoding dependency, for example, using inter-channel prediction), and to encode the N channels into the data stream 12 with a mutual delay 35 according to the inter-channel delay information 23.

[0185] The multi-channel digital signal 10 may have N channels, and the encoder 200 may be configured to encode the N channels into the data stream 12 in a manner that delays each other by 35 according to the inter-channel delay information 23, such that the samples of two channels in the K channels 20 are subject to a mutual delay 35 greater than one sampling interval.

[0186] 2.6 Bitstream permutations for low-latency coding

[0187] What might be of interest is arranging the bitstream in a way that (e.g., bitstream 12) such that only up to a predefined time is needed. Sample values ​​(e.g., measured in time and / or samples) are used to generate decodeable portions of the bitstream in order to reconstruct the sequence. (e.g., a multi-channel digital signal 10) up to a predefined time The sample. Difference (For example, a time interval of 26) represents the encoding delay associated with that decodeable portion. In a communication scenario, it might be of interest to arrange the bitstream in a manner that ensures no bitstream portion exceeds a specific encoding delay.

[0188] Suppose that the sub-bitstreams generated by the target encoder can be arranged into parts that can be decoded sequentially (e.g., sub-stream part 30), such that each decoding part can produce some reconstructed sample values, which can be used to reconstruct (potentially distorted) sequences. Use when needed.

[0189] In this invention, the bitstream can be arranged into ordered portions (e.g., substream portion 30) that can be decoded in a specific order. The decoder can decode a new portion whenever the encoder generates it. Let... For the bit stream Each part is set up To generate part Required sequence At the time of the sample value, let To pass the decoding part Can be refactored The moment of the (potentially distorted) sample value.

[0190] In a preferred embodiment, each bitstream portion Arranged in a way that results in associated delays (For example, a time interval of 26 or a delay based thereon) does not exceed the predetermined maximum delay value.

[0191] Figure 11a It shows Figure 1a (or Figure 5a The decoder 100, wherein the biophysiological signal 102 (or any other type of signal disclosed herein, such as an audio signal or seismic data) is a multi-channel digital signal 10, and decoding the multi-channel digital signal 10 from the data stream 12 using an audio decoding scheme (or a decoding scheme with similar functionality) includes: grouping 21 the K' coded channels 14 representing the multi-channel digital signal 10 into Q sets 25, each set having n iThere are n encoding channels, where n i Indicates the number of encoded channels in set i, Q ≥ i ≥ 1 (e.g., based on channel grouping information 40 decoded from data stream 12), and uses an audio decoding scheme to decode at least one set j from Q sets 25 (for set n) of data stream 12. j ≥2), where the audio decoding scheme is a multi-channel audio decoding scheme, involving: decoding m-channel downmixed signal 32 from data stream 12 (where m < n) j ), and by using auxiliary information 34 contained in data stream 12 to upmix the m-channel downmixed signal 32, the n of set j is obtained. j 14 encoding channels.

[0192] Decoder 100 can be further configured to decode each of the Q sets 25 from the substream 28 associated with the corresponding set 25 in the data stream 12 at time intervals 26 (e.g., intervals 26a, b, c), such that each substream 28 is formed by a sequence of substream portions 30 (where in Figure 11a In the data stream 12, only the substream portions 30 marked with "1" are depicted with reference numeral 30 to better visualize their association with the same set, but the substream portions 30 marked with "2" are also substream portions 30. The sequence of substream portions has consecutive time intervals encoded in it in the set 25 associated with the corresponding substream 28, and the substream portions 30 of substream 28 are interleaved (e.g., encoded in an alternating manner) such that substream portions 30 whose time intervals overlap are adjacent and consecutive in the data stream 12, and precede the substream portion 30 whose time interval is temporally subsequent.

[0193] Similarly, Figure 5a The decoder can be configured to decode each of Q sets 25 from substreams 28 of data stream 12 associated with corresponding sets 25 in units of time intervals 26, wherein each substream 28 is formed by a sequence of substream portions 30 having consecutive time intervals 26 encoded therein in the set 25 associated with the corresponding substream 28, and the substream portions 30 of the substreams are interleaved such that substream portions 30 with overlapping time intervals 26 are adjacent and consecutive in data stream 12, and that the substream portion 30 precedes its time interval 26 in time. However, the same principle can be applied to any other decoder 100 disclosed herein.

[0194] Figure 11b It shows Figure 1b (or Figure 5bThe encoder 200, wherein the biophysiological signal (or any other type of signal disclosed herein, such as an audio signal or seismic data) is a multi-channel digital signal 10, and encoding the multi-channel digital signal 10 into a data stream 12 using an audio coding scheme includes: grouping K' coded channels 14 representing the multi-channel digital signal 10 into Q sets 25, each set having n i There are 14 encoding channels, where n i The number of encoded channels 14 in set i is indicated, Q ≥ i ≥ 1 (e.g., where the grouping can be derived based on the channel grouping information 40 encoded into data stream 12), and at least one set j of the Q sets (for the set n) is encoded using audio coding scheme 83. j ≥2) Encode into data stream 12, where the audio encoding scheme is a multi-channel audio encoding scheme, involving: downmixing m channels of signal 32 (where m < n) j The signal is encoded into data stream 12 so that the set j can be obtained by upmixing the m-channel downmixed signal 32 using auxiliary information 34 contained in the data stream. j 14 encoding channels.

[0195] The encoder 200 can be configured to encode each of the Q sets 25 into a substream 28 of the data stream 23 associated with the corresponding set in units of time interval 26, such that each substream 28 is formed by a sequence of substream portions 30 having consecutive time intervals 26 encoded therein in the set 25 associated with the corresponding substream 28, and the substream portions 30 of the substream 28 are interleaved such that substream portions 30 whose time intervals 26 overlap are adjacent and consecutive in the data stream 12, and that the substream portions 30 that precede their time intervals 26 in time are the subsequent substream portions 30.

[0196] Similarly, Figure 5b The decoder can be configured to encode each of the Q sets 25 into a substream 28 of data stream 12 associated with the corresponding set 25 in units of time intervals 26, wherein each substream 28 is formed by a sequence of substream portions 30 having consecutive time intervals 26 encoded therein in sets associated with the corresponding substream 28, and the substream portions 30 of substream 28 are interleaved such that substream portions 30 whose time intervals 26 overlap are adjacent and consecutive in data stream 12, and that the substream portions 30 that precede their time intervals 26 in time are the subsequent substream portions 30. However, the same principle can be applied to any other encoder 200 disclosed herein.

[0197] The time interval 26 can have the same length, for example, in terms of sample size or time duration. The time interval 26 can be (or based on) Figure 4The sample length of time block 65 (e.g., with 16 samples 63, or any other number of samples 63).

[0198] 2.7 Bitstream arrangement for random access decoding

[0199] What may be of interest is arranging the bitstream in a way that supports random access decoding. That is, the decoder (e.g., decoder 100) can identify the positions in the bitstream that support random access decoding and start decoding from there.

[0200] Assuming any part of the sub-bitstream (e.g.) Figure 4 The time block 65 shown is, for example, by Figure 4 The boundary 96 and / or the boundary separator next to the access encoding channels 88a and 88b can be arranged in a way that allows decoding to begin from there.

[0201] In this invention, an encoder (e.g., encoder 200) can generate sub-bit streams in a manner using a target encoder such that they can be arranged in portions, and some of these portions can be used to begin decoding from them.

[0202] In a preferred embodiment, the bitstream is arranged into sections (e.g., time blocks 65), some of which support random access decoding (e.g., time blocks 65 separated by boundary 96).

[0203] In another preferred embodiment, the bitstream is arranged in a manner that allows the following decoding process to be employed. The decoder's random access position within the bitstream (e.g., Figure 4 The decoder begins decoding the bitstream at time blocks 65b and 65e. It extracts portions of the sub-bitstream and identifies the random access positions contained within them. It then starts decoding the sub-bitstream from those random access positions. This produces decoded sample sequences for different channels, which may begin at different times. That is, for some times, there may be decoded samples for some channels, while other channels may not have decoded samples. Therefore, the decoder identifies the first moment when decoded samples have been produced for all channels. This is the moment when random access decoding begins.

[0204] In another preferred embodiment, the bitstream portion supporting random access also allows for zero-level, one-level, or multi-level linear combinations and / or optional subsequent inter-channel delay adjustments when reconstructing the signal.

[0205] For example, for Figure 11aIn the decoder 100, some substream portions of the sequence of substream portions 30 in each substream 28 can be encoded as random access points. However, the same principle can be applied to any other decoder 100 disclosed herein (e.g., decoder 100 configured to decode each of Q sets from substream 28 of data stream 12).

[0206] Random access points in substream 28 can be time-aligned. Alternatively, random access points in substream 28 can be non-time-aligned.

[0207] In each substream 28, some substream portions of the sequence of substream portion 30 can be encoded as random access points (e.g., with...). Figure 4 The sub-stream portion 30 related to time blocks 65b and 65e in the middle.

[0208] Similarly, for Figure 11b In the encoder 200, some substream portions of the sequence of substream portions 30 in each substream 28 can be encoded as random access points. However, the same principle can be applied to any other encoder 200 disclosed herein (e.g., an encoder 200 configured to encode each of Q sets 25 into substream 28 of data stream 12).

[0209] 2.8 Supports channels with different sampling rates and / or types.

[0210] As mentioned earlier, in In some instances (e.g., multi-channel digital signal 10), different sampling rates may be used in different channel signals (e.g., channels 14 and / or 20). Furthermore, some signals may be bandwidth-limited to such an extent that encoding in a subsampled representation and transmitting in a bitstream may be feasible without significantly increasing distortion during encoding. Finally, some channel signals may not actually be biomedical signals, but rather human-readable annotated text; these documents may be sparse and temporally related to… Alignment with certain characteristics of one or more other (biomedical) channel signals, or with some or even other types of non-waveform signal data. Similarly, the multi-channel digital signals 10 described herein, which include different types of data (e.g., seismic measurement data or weather data), can also be used.

[0211] In response to Different possible characteristics of the signal in the middle, it is recommended to refer to The individual signal characteristics of the mid-channel signal allow for the use of different coding techniques, or in other words, different codec specifications. Therefore, a preferred embodiment (e.g., primarily focused on encoding but similarly applicable to decoding) may be as follows.

[0212] 1) Typical biomedical (or other data source) waveform data: According to any of the previously described embodiments, the corresponding channel signals of this type (e.g., channel 20) are clustered (e.g., grouped according to channels that have or do not have periodicity and / or waveform type characteristics, such as grouping based on one or more criteria and / or predefined groups, such as according to data source input), and then waveform encoding techniques (time domain and / or transform domain) are used, such as AAC, HE-AAC (High Efficiency Advanced Audio Coding), xHE-AAC, or other lossless or lossy waveform encoders.

[0213] 2) Band-limited waveform data: Clustering of the corresponding channel signals of this type according to any of the previously described embodiments (e.g., Figure 3 The signal is grouped into groups 44 (e.g., based on similarity of one or more features including amplitude, frequency, spectrum, and offset), then downsampled (e.g., by a factor of 2 or any other factor) and waveform encoding is applied. To undo the downsampling, the decoded channel signal is upsampled (e.g., by a factor of 2) after being decoded by the waveform encoder employed. Note that the waveform encoder does not necessarily have to be the same as the encoder used in case 1). In fact, when the same codec is used on the downsampled channel signal, the frames of that codec may span a longer time period (e.g., when using an AAC encoder variant and downsampling by 2, the time period becomes twice as long). To synchronize framing between the encoding of band-limited and unband-limited channel signals, it may be necessary to use a waveform encoder with a shorter (and preferably proportionally) frame size than the waveform encoder used in case 1) in case 2). To give a general example, not intended to limit the scope of this embodiment to the AAC encoder family: AAC-LD uses a shorter frame size than AAC or (x)HE-AAC.

[0214] 3) Text data, such as human-readable data: This type of data (which is usually sparse in time) can be most efficiently encoded by, for example, run-length encoding and / or “zip” techniques (such as Lempel-Ziv or Lempel-Ziv-Welch algorithms), preferably but not necessarily in a lossless manner, as opposed to an encoder oriented towards that waveform.

[0215] As an implementation of methods and data streams

[0216] Furthermore, a method performed by any decoder 100 disclosed herein is provided. For example, a method for decoding a biophysiological signal 102 from a data stream 12 is provided, the method comprising decoding the biophysiological signal 102 from the data stream 12 using an audio decoding scheme (e.g., which may be derived by...). Figure 1a The decoder 100 shown is executed by it. Similarly, methods that can be executed by any decoder 100 disclosed herein are also provided, such as Figure 2 , 4 The decoders shown are 5a, 6a, 7a, 8a, 9a, 10a and 11a.

[0217] Furthermore, a method performed by any encoder 100 disclosed herein is provided. For example, a method for encoding a biophysiological signal 102 into a data stream 12 is provided, the method comprising encoding the biophysiological signal 102 into the data stream 12 using an audio coding scheme (e.g., which may be generated by an encoder 100). Figure 1b The decoder 100 shown is executed by this. Similarly, methods that can be executed by any encoder 200 disclosed herein are provided, such as... Figure 3 , 4 The encoders shown are 5b, 6b, 7b, 8b, 9b, 10b, and 11b.

[0218] Furthermore, a bitstream 12 (or data stream or data stream including bitstream 12) encoded by any encoder 200 (or method performed by such encoder 200) disclosed herein is provided. Bitstream 12 can be decoded by any decoder 100 (or method performed by such decoder 100) disclosed herein. Bitstream 12 can be stored on a digital storage medium, such as a non-transitory storage medium.

[0219] Furthermore, other embodiments will be defined by the appended claims.

[0220] It should be noted that any embodiment defined in the claims may be supplemented by any details (features and functions) described in the foregoing sections.

[0221] Furthermore, the embodiments described in the foregoing sections can be used alone or supplemented by any feature in another section or any feature included in the claims.

[0222] Furthermore, it should be noted that the various aspects described herein can be used individually or in combination. Therefore, details can be added to each of these aspects without needing to add details to the other aspects.

[0223] Furthermore, the features and functions disclosed herein related to the methods can also be used in apparatuses (configured to perform such functions). Additionally, any features and functions disclosed herein regarding the apparatus can also be used in the corresponding methods. In other words, the methods disclosed herein can be supplemented by any features and functions described regarding the apparatus.

[0224] Furthermore, any features and functions described herein can be implemented in hardware or software, or a combination of hardware and software, as described in the "Implementation Alternatives" section.

[0225] Furthermore, any features and functions described herein can be implemented in hardware or software, or a combination of hardware and software, as described in the "Implementation Alternatives" section.

[0226] Implement an alternative solution:

[0227] Although some aspects have already been described in the context of the apparatus (e.g., an encoder), these aspects also clearly represent a description of the corresponding method, where blocks or devices correspond to method steps or features of method steps. Similarly, aspects described in the context of method steps also represent a description of corresponding blocks, items, or features of the corresponding apparatus. Some or all of the method steps may be performed by (or using) hardware devices, such as microprocessors, programmable computers, or electronic circuits. In some embodiments, one or more of the most important method steps may be performed by such devices.

[0228] Depending on certain implementation requirements, embodiments of the present invention can be implemented in hardware or software. The implementation can be performed using digital storage media, such as floppy disks, DVDs, Blu-ray discs, CDs, ROMs, PROMs, EPROMs, EEPROMs, or flash memory, storing electronically readable control signals that cooperate (or are capable of cooperating with) a programmable computer system to perform corresponding methods. Therefore, the digital storage medium can be computer-readable. The encoded media signals can be encoded into a data stream. The data stream can be stored on the aforementioned digital storage medium (e.g., a temporary digital storage medium).

[0229] Some embodiments of the invention include a data carrier having electronically readable control signals that are capable of cooperating with a programmable computer system to perform one of the methods described herein.

[0230] Typically, embodiments of the present invention can be implemented as a computer program product having program code that, when run on a computer, performs one of the methods. The program code may, for example, be stored on a machine-readable medium (e.g., a non-transitory storage medium).

[0231] Other embodiments include a computer program for performing one of the methods described herein, stored on a machine-readable medium.

[0232] Therefore, in other words, an embodiment of the method of the present invention is a computer program having program code that, when run on a computer, performs one of the methods described herein.

[0233] Other embodiments of the method of the present invention are therefore data carriers (or digital storage media, or computer-readable media) having a computer program recorded thereon for performing one of the methods described herein. Data carriers, digital storage media, or recording media are generally tangible and / or non-transitory.

[0234] Therefore, other embodiments of the method of the present invention represent data streams or signal sequences for performing one of the methods described herein. The data streams or signal sequences may, for example, be configured to be transmitted via a data communication connection (e.g., via the Internet).

[0235] Other embodiments include processing means, such as a computer or programmable logic device, configured or adapted to perform one of the methods described herein.

[0236] Other embodiments include a computer on which a computer program for performing one of the methods described herein is installed.

[0237] Other embodiments of the invention include an apparatus or system configured to transmit (e.g., electronically or optically) a computer program for performing one of the methods described herein to a receiver. The receiver may be, for example, a computer, a mobile device, a storage device, etc. The apparatus or system may include, for example, a file server for transmitting the computer program to the receiver.

[0238] In some embodiments, a programmable logic device (e.g., a field-programmable gate array) may be used to perform some or all of the functions of the methods described herein. In some embodiments, the field-programmable gate array may cooperate with a microprocessor to perform one of the methods described herein. Generally, these methods are preferably performed by any hardware device.

[0239] The apparatus described herein can be implemented using hardware devices, computers, or a combination of hardware devices and computers.

[0240] The apparatus or any component thereof described herein may be implemented, at least in part, in hardware and / or software.

[0241] The methods described herein can be performed using hardware devices, computers, or a combination of hardware devices and computers.

[0242] The methods described herein, or any component thereof, may be executed, at least in part, by hardware and / or software.

[0243] The above embodiments are merely illustrative of the principles of the present invention. It should be understood that modifications and variations of the arrangements and details described herein will be apparent to those skilled in the art. Therefore, the intent is limited only by the scope of the forthcoming patent claims, and not by the specific details of the embodiments described and explained herein.

Claims

1. A decoder (100) for decoding biophysiological signals (102) from a data stream (104; 12), configured to: The biophysiological signal (102) is decoded from the data stream (104; 12) using an audio decoding scheme (106).

2. The decoder (100) according to claim 1. The audio decoding scheme (106) mentioned above involves: Transform-based audio decoding, wherein the transform-based audio decoding includes: The scaling factor and transformation coefficient are derived from the data stream (12), and the transformation coefficient is spectrally shaped using the scaling factor to obtain the shaped spectrum. The shaped spectrum is then transformed again to obtain the audio frame signal, and the audio frame signal is subjected to overlapping addition processing. or Linear predictive coding (LPC) audio decoding includes: deriving LPC coefficients and information about the residual signal from the data stream, and using the LPC coefficients to subject the residual signal to LPC synthesis.

3. The decoder (100) according to claim 2, wherein the audio decoding scheme (106) relates to: The scaling factor is derived from the data stream (12), which involves: entropy decoding the scaling factor from the data stream (12), or decoding LPC coefficients from the data stream (12) and converting the LPC coefficients into the scaling factor.

4. The decoder (100) according to any one of claims 1 to 3, wherein: The biophysiological signal is a multi-channel digital signal (10), and Decoding the multi-channel digital signal (10) from the data stream (12) using the audio decoding scheme (106) includes: decoding K' encoded channels (14) representing the multi-channel digital signal (10) from the data stream (12) in the following manner: The audio decoding scheme (106) is used to decode n > 1 encoded channels from the data stream (12), where n ≤ K', wherein the audio decoding scheme (106) is a multi-channel audio decoding scheme (29), which involves: decoding (31) m-channel downmixed signal (32) from the data stream (12), where 0 < m < n, and upmixing (33) the m-channel downmixed signal using auxiliary information (34) contained in the data stream (12) to obtain n encoded channels.

5. The decoder (100) according to any one of claims 1 to 4, wherein: The biophysiological signal is a multi-channel digital signal (10), and Decoding the multi-channel digital signal (10) from the data stream (12) using the audio decoding scheme includes: The K' encoded channels (14) representing the multi-channel digital signal (10) are grouped into Q sets, each set having n i There are n encoding channels, where n i Indicates the number of encoded channels in set i, where Q ≥ i ≥ 1, and The audio decoding scheme is used to decode at least one set j of the Q sets from the data stream (12), where n j ≥2, the audio decoding scheme is a multi-channel audio decoding scheme, which involves: decoding m-channel downmixed signals (32) from the data stream (12), where m < n j The set j is obtained by upmixing the m-channel downmixed signal using the auxiliary information (34) contained in the data stream (12). j One encoding channel (14).

6. The decoder (100) according to claim 4 or 5 is configured to: The K' encoded channels (14) of the multi-channel digital signal (10) are subjected to channel retransformation, which retransforms the co-aligned sample positions (18) of the K' encoded channels (14) to obtain the K channels (20) of the multi-channel digital signal (10).

7. The decoder (100) according to claim 6 is configured to derive the channel retransformation (16) from the transformation information (22) in the data stream (12).

8. The decoder (100) according to claim 7 is configured to update the channel retransformation (16) based on the transformation information (22) in the data stream (12).

9. The decoder (100) according to any one of claims 6 to 8, wherein the channel retransformation (16) involves a sequence of partial channel retransformations (24).

10. The decoder (100) according to claim 9, wherein: The partial channel retransformation has different dimensions in the number of retransformed channels, so that different encoded channels are affected by different subsets of transformations, different numbers of transformations, or different subsequences of transformations in the sequence of partial channel retransformation.

11. The decoder (100) according to claim 5 or any one of claims 6 to 10, wherein: The multi-channel decoding scheme is designed to handle a maximum of MAX encoded channels to the greatest extent possible, and The decoder is configured to group the K' encoded channels (14) representing the multi-channel digital signal (10) into Q sets, each set having n i n encoding channels, such that for each 0 < i < Q+1, n i ≤MAX.

12. The decoder (100) according to any one of claims 5 to 11 is configured as follows: Decode channel grouping information (40) from the data stream (12) and perform grouping using the channel grouping information.

13. The decoder (100) according to claim 12, wherein: The channel grouping information (40) includes: Indicates the number n of the encoded channels in the Q sets. i syntax elements, and / or The sequential channel order among the K' encoded channels (14) is used to group the K' encoded channels (14) into the Q sets such that each encoded channel of different sets in the Q sets is non-interleaved along the sequential channel order.

14. The decoder (100) according to any one of claims 4 to 13, wherein the auxiliary information includes one or more of the following: Inter-channel coherence ICC data, Channel level difference (CLD) data, and Channel prediction coefficients (CPC) data.

15. The decoder (100) according to any one of claims 4 to 14, wherein the decoder is configured to: Decode the inter-channel delay information (23) from the data stream (12), and The channels (20) of the multi-channel digital signal (10) are delayed to each other according to the inter-channel delay information (23).

16. The decoder (100) according to any one of claims 4 to 15 is configured to: The channels of the multi-channel digital signal (10) are replaced (50) according to the channel replacement information (52) notified in the data stream (12) so as to obtain the multi-channel digital signal (10).

17. The decoder (100) according to claim 16, wherein: The channel permutation information (52) describes the permutation of the channels of the multi-channel digital signal (10) obtained by the channel retransformation (16) in order to obtain a predetermined representation of the multi-channel digital signal (10).

18. The decoder (100) according to any one of claims 5 to 17 is configured to: Decode each of the Q sets from the substream (28) associated with the corresponding set of the data stream in units of time intervals (26), such that: Each substream is formed by a sequence of substream portions (30), the sequence of which has consecutive time intervals encoded therein as associated with a set of corresponding substreams, and The substream portions of the substream are interleaved such that substream portions with overlapping time intervals are adjacent and continuous in the data stream, and that the substream portions with overlapping time intervals precede the substream portions with overlapping time intervals.

19. The decoder (100) according to claim 18, wherein: In each substream (28), some substream portions of the sequence of substream portions (30) are encoded as random access points.

20. The decoder (100) according to claim 19, wherein: The random access points of the substreams are time-aligned.

21. The decoder (100) according to claim 19, wherein: The random access points of the substream are not aligned in time.

22. An encoder (200) for encoding biological physiological signals (102) into a data stream (104), configured to: The biophysiological signals are encoded into the data stream using an audio coding scheme (206).

23. The encoder (200) according to claim 22. The audio encoding scheme (206) mentioned therein involves: Transform-based audio coding, wherein the transform-based audio coding includes: The scaling factor and transformation coefficient are inserted into the data stream (12), wherein the scaling factor is used to perform spectral shaping on the transformation coefficient to obtain a shaped spectrum, the shaped spectrum is re-transformed to obtain an audio frame signal, and the audio frame signal is subjected to overlapping and addition processing to generate reconstruction. or Linear predictive coding (LPC) audio coding includes inserting LPC coefficients and information about the residual signal into the data stream (12) such that the residual signal is subjected to LPC synthesis using the LPC coefficients to generate a reconstruction.

24. The encoder (200) according to claim 23, wherein the audio encoding scheme relates to: Inserting the scaling factor into the data stream (12) involves either encoding the scaling factor entropy into the data stream (12) or encoding the LPC coefficients into the data stream (12) to allow the LPC coefficients to be converted into the scaling factor.

25. The encoder (200) according to any one of claims 22 to 24, wherein: The biophysiological signal is a multi-channel digital signal (10), and Encoding the multi-channel digital signal (10) into the data stream (12) using the audio encoding scheme includes encoding K' encoding channels (14) representing the multi-channel digital signal (10) into the data stream (12) in the following manner: The audio encoding scheme is used to encode n>1 encoding channels into the data stream (12), where n≤K'. The audio encoding scheme is a multi-channel audio encoding scheme (29), which involves: encoding m-channel upmixed signals (32) into the data stream (12), where 0<m<n, and downmixing the m-channel upmixed signals (33) using auxiliary information (34) encoded into the data stream (12), so that n encoding channels can be obtained from the data stream (12).

26. The encoder (200) according to any one of claims 22 to 25, wherein: The biophysiological signal is a multi-channel digital signal (10), and Encoding the multi-channel digital signal (10) into the data stream (12) using the audio encoding scheme includes: The K' encoded channels (14) representing the multi-channel digital signal (10) are grouped into Q sets, each set having n i There are n encoding channels, where n i Indicates the number of encoded channels in set i, where Q ≥ i ≥ 1, and The audio encoding scheme is used to encode at least one set j of the Q sets into the data stream (12), where n j ≥2, the audio encoding scheme is a multi-channel audio encoding scheme, which involves encoding m channels of downmixed signal (32) into the data stream (12), where m < n j This allows the set j to be obtained by upmixing the m-channel downmixed signal using the auxiliary information (34) contained in the data stream (12). j One encoding channel (14).

27. The encoder (200) according to claim 25 or 26 is configured to: The K channels (20) of the multi-channel digital signal (10) are subjected to channel transformation, which transforms the co-aligned sample positions of the K channels to obtain the K' encoded channels (14) of the multi-channel digital signal (10).

28. The encoder (200) according to claim 27 is configured to include transformation information (22) in the data stream (12), from which a channel retransformation (16) corresponding to the channel transformation can be derived.

29. The encoder (200) according to claim 28 is configured to change the channel transformation and accordingly signal the update of the transformation information in the data stream (12).

30. The encoder (200) according to any one of claims 27 to 29, wherein the channel transformation involves a sequence of partial channel transformations (24).

31. The encoder (200) according to claim 30, wherein: The partial channel transformation (24) has different dimensions in the number of transformation channels, so that different coding channels are affected by different transformation subsets, different transformation numbers or different transformation subsequences in the sequence of the partial channel transformation (24).

32. The encoder (200) according to any one of claim 24 or claim 31, wherein: The multi-channel encoding scheme is designed to handle a maximum of MAX encoding channels to the greatest extent possible, and The encoder is configured to group the K' encoded channels (14) representing the multi-channel digital signal (10) into Q sets, each set having n i n encoding channels, such that for each 0 < i < Q+1, n i ≤MAX.

33. The encoder (200) according to any one of claims 26 to 32 is configured as follows: The channel grouping information (40) indicating the grouping is encoded into the data stream (12).

34. The encoder (200) according to claim 33, wherein, The channel grouping information (40) includes: Indicates the number n of the encoded channels in the Q sets. i syntax elements, and / or The sequential channel order among the K' encoded channels (14) is used to group the K' encoded channels (14) into the Q sets, such that each encoded channel of different sets in the Q sets is non-interleaved along the sequential channel order.

35. The encoder (200) according to any one of claims 25 to 34, wherein the auxiliary information includes one or more of the following: Inter-channel coherence ICC data, Channel level difference (CLD) data, and Channel prediction coefficients (CPC) data.

36. The encoder (200) according to any one of claims 25 to 35, wherein the encoder is configured to: The inter-channel delay information (23) is encoded into the data stream (12), and the K channels (20) are encoded into the data stream (12) while the K channels are mutually delayed (35) according to the inter-channel delay information (23).

37. The encoder (200) according to any one of claims 25 to 36 is configured to: Includes channel permutation information (52) of the signal notification in the data stream (12), and permutes the channels of the multi-channel digital signal (10) so that the channels of the multi-channel digital signal (10) are encoded in the permuted state.

38. The encoder (200) according to claim 37, wherein: The channel permutation information (52) describes the permutation between the channels of the multi-channel digital signal (10) encoded into the data stream (12) and a predetermined representation of the multi-channel digital signal (10).

39. The encoder (200) according to any one of claims 26 to 38 is configured to: Each of the Q sets is encoded into a substream (28) of the data stream associated with the corresponding set at time intervals (26), such that: Each substream is formed by a sequence of substream portions (30), the sequence of which has consecutive time intervals encoded therein as associated with a set of corresponding substreams, and The sub-stream portions of the sub-stream are interleaved such that sub-stream portions with overlapping time intervals are adjacent and continuous in the data stream, and precede the sub-stream portions whose time intervals are temporally subsequent.

40. The encoder (200) according to claim 39, wherein: In each substream, some substream portions of the sequence of substream portions are encoded as random access points.

41. The encoder (200) according to claim 40, wherein: The random access points of the substreams are time-aligned.

42. The encoder (200) according to claim 40, wherein: The random access points of the substream are not aligned in time.

43. A decoder (100) for decoding a multi-channel digital signal (10) from a data stream (12), configured to: Decode channel packet information (40) from the data stream (12), and Based on the channel grouping information, the K' encoded channels (14) representing the multi-channel digital signal (10) are grouped (21) into Q sets, each set having n i There are n encoding channels, where n i Indicates the number of encoded channels in set i, where Q ≥ i ≥ 1, and A multi-channel decoding scheme is used to decode at least one set j of the Q sets from the data stream (12), where n j ≥2.

44. The decoder (100) according to claim 43, wherein: The channel grouping information (40) includes: Indicates the number n of the encoded channels in the Q sets. i syntax elements, and / or The sequential channel order among the K' encoded channels (14) is used to group the K' encoded channels (14) into the Q sets such that the encoded channels of different sets in the Q sets are not interleaved along the sequential channel order, and one or more encoded channels in a set are adjacent to each other in the sequential channel order.

45. The decoder (100) according to claim 43 or 44 is configured to: The multi-channel decoding scheme is used to decode each of the Q sets j from the data stream (12), where, n j ≥2。 46. ​​The decoder (100) according to any one of claims 43 to 45, wherein: The multi-channel decoding scheme involves decoding m-channel downmixed signals from the data stream (12), where m < n. j The n signal is obtained by upmixing the m-channel downmixed signal using the auxiliary information contained in the data stream (12). j n encoding channels, to obtain the n j One encoding channel.

47. The decoder (100) according to claim 46, wherein the auxiliary information includes one or more of the following: Inter-channel coherence ICC data, Channel level difference (CLD) data, and Channel prediction coefficients (CPC) data.

48. The decoder (100) according to any one of claims 43 to 47, wherein: The multi-channel decoding scheme is designed to handle a maximum of MAX encoded channels to the greatest extent possible, and The decoder is configured to group the K' encoded channels (14) representing the multi-channel digital signal (10) into Q sets, each set having n i n encoding channels, such that for each 0 < i < Q+1, n i ≤MAX.

49. The decoder (100) according to any one of claims 43 to 48 is configured to: The K' encoded channels (14) representing the multi-channel digital signal (10) are subjected to channel retransformation (16), which retransforms the co-aligned sample positions (18) of the K' encoded channels (14) to obtain the K channels of the multi-channel digital signal (10).

50. The decoder (100) according to claim 49 is configured to derive the channel retransformation (16) from the transformation information (22) in the data stream (12).

51. The decoder (100) according to claim 50 is configured to update the channel retransformation (16) based on the transformation information (22) in the data stream (12).

52. The decoder (100) according to any one of claims 49 to 51, wherein the channel retransformation (16) involves a sequence of partially channel retransformations.

53. The decoder (100) according to claim 52, wherein: The partial channel retransformation has different dimensions in the number of retransformed channels, so that different encoded channels are affected by different subsets of transformations, different numbers of transformations, or different subsequences of transformations in the sequence of partial channel retransformation.

54. The decoder (100) according to any one of claims 43 to 53, wherein: The multi-channel decoding scheme is a multi-channel audio decoding scheme.

55. The decoder (100) according to any one of claims 43 to 54. The multi-channel decoding scheme mentioned above involves: Transform-based decoding, which includes: The scaling factor and transformation coefficient are derived from the data stream (12), and the transformation coefficient is spectrally shaped using the scaling factor to obtain a shaped spectrum. The shaped spectrum is then re-transformed to obtain a frame signal, and the frame signal is subjected to overlapping addition processing. or Linear predictive coding (LPC) decoding, wherein the LPC decoding includes: deriving LPC coefficients and information about the residual signal from the data stream, and using the LPC coefficients to subject the residual signal to LPC synthesis.

56. The decoder (100) according to claim 55, wherein: Derivation of the scaling factor from the data stream (12) involves: entropy decoding of the scaling factor from the data stream (12), or decoding of LPC coefficients from the data stream (12) and converting the LPC coefficients into the scaling factor.

57. The decoder (100) according to any one of claims 43 to 56, wherein the multi-channel digital signal (10) has K channels, and the decoder is configured to: Decode the inter-channel delay information (23) from the data stream (12), and The K channels are delayed to each other according to the inter-channel delay information (23).

58. The decoder (100) according to any one of claims 43 to 57. The multi-channel digital signal (10) mentioned therein includes biological and physiological signals.

59. The decoder (100) according to any one of claims 43 to 58 is configured to: The channels of the multi-channel digital signal (10) are replaced (50) according to the channel replacement information (52) notified in the data stream (12) so as to obtain the multi-channel digital signal (10).

60. The decoder (100) according to claim 59, wherein: The channel permutation information (52) describes the permutation of the channels in order to obtain a predetermined representation of the multi-channel digital signal (10).

61. The decoder (100) according to any one of claims 43 to 60 is configured to: Decode each of the Q sets from the substream (28) associated with the corresponding set of the data stream in units of time intervals (26), where: Each substream is formed by a sequence of substream portions (30), the sequence of which has consecutive time intervals encoded therein as associated with a set of corresponding substreams, and The substream portions of the substream are interleaved such that substream portions with overlapping time intervals are adjacent and continuous in the data stream, and that the substream portions with overlapping time intervals precede the substream portions with overlapping time intervals.

62. The decoder (100) according to claim 61, wherein: In each substream, some substream portions of the sequence of substream portions are encoded as random access points.

63. The decoder (100) according to claim 62, wherein: The random access points of the substreams are time-aligned.

64. The decoder (100) according to claim 62, wherein: The random access points of the substream are not aligned in time.

65. An encoder (200) for encoding a multi-channel digital signal (10) into a data stream (12), configured to: The channel grouping information is encoded into the data stream (12). Using the channel grouping information, it is indicated that the K' encoded channels (14) of the multi-channel digital signal (10) are grouped into Q sets, each set having n i There are n encoding channels, where n i Indicates the number of encoded channels in set i, where Q ≥ i ≥ 1, and At least one set j from the Q sets is encoded into the data stream (12) using a multi-channel encoding scheme, where n j ≥2.

66. The encoder (200) according to claim 53, wherein: The channel grouping information includes: Indicates the number n of the encoded channels in the Q sets. i syntax elements, and / or The sequential channel order among the K' encoded channels (14) is used to group the K' encoded channels (14) into the Q sets such that the encoded channels of different sets in the Q sets are not interleaved along the sequential channel order, and one or more encoded channels in a set are adjacent to each other in the sequential channel order.

67. The encoder (200) according to claim 53 or 54 is configured to: The multi-channel encoding scheme is used to encode each set j of the Q sets into the data stream (12), where n j ≥2.

68. The encoder (200) according to any one of claims 53 to 55, wherein: The multi-channel encoding scheme involves encoding m channels of downmixed signal into the data stream (12), where m < n. j By upmixing the m-channel downmixed signal using the auxiliary information contained in the data stream (12), n can be obtained from the data stream (12). j One encoding channel.

69. The encoder (200) according to claim 56, wherein the auxiliary information includes one or more of the following: Inter-channel coherence ICC data, Channel level difference (CLD) data, and Channel prediction coefficients (CPC) data.

70. The encoder (200) according to any one of claims 53 to 57, wherein: The multi-channel encoding scheme is designed to handle a maximum of MAX encoding channels to the greatest extent possible, and The encoder is configured to group the K' encoded channels (14) representing the multi-channel digital signal (10) into Q sets, each set having n i n encoding channels, such that for each 0 < i < Q+1, n i ≤MAX.

71. The encoder (200) according to any one of claims 43 to 58 is configured as follows: The K channels of the multi-channel digital signal (10) are subjected to channel transformation, which transforms the co-aligned sample positions of the K channels to obtain K' encoded channels (14) of the multi-channel digital signal (10).

72. The encoder (200) of claim 59 is configured to include transformation information (22) in the data stream (12) to allow a channel retransformation (16) corresponding to the channel transformation to be derived.

73. The encoder (200) according to claim 60 is configured to signal updates to the transformation information (22) in the data stream (12) in order to update the channel transformation.

74. The encoder (200) according to any one of claims 59 to 60, wherein the channel transformation involves a sequence of partial channel transformations (24).

75. The encoder (200) according to claim 62, wherein: The partial channel transformation (24) has different dimensions in the number of transformation channels, so that different coding channels are affected by different transformation subsets, different transformation numbers or different transformation subsequences in the sequence of the partial channel transformation (24).

76. The encoder (200) according to any one of claims 53 to 63, wherein: The multi-channel coding scheme is a multi-channel audio coding scheme.

77. The encoder (200) according to any one of claims 53 to 64. The multi-channel coding scheme mentioned above involves: Transform-based coding, wherein the transform-based coding includes: The scaling factor and transformation coefficients are inserted into the data stream (12), wherein the scaling factor is used to perform spectral shaping on the transformation coefficients to obtain a shaped spectrum, the shaped spectrum is re-transformed to obtain a frame signal, and the frame signal is subjected to overlapping and addition processing to generate reconstruction; or Linear predictive coding (LPC) encoding, wherein the LPC encoding includes inserting LPC coefficients and information about the residual signal into the data stream, such that the residual signal is subjected to LPC synthesis using the LPC coefficients to generate a reconstruction.

78. The encoder (200) according to claim 65, wherein: Inserting the scaling factor into the data stream (12) involves either encoding the scaling factor entropy into the data stream (12) or encoding the LPC coefficients into the data stream (12) to allow the LPC coefficients to be converted into the scaling factor.

79. The encoder (200) according to any one of claims 53 to 66, wherein the multi-channel digital signal (10) has K channels, and the encoder is configured to: The inter-channel delay information (23) is encoded into the data stream (12), and While the K channels are mutually delayed according to the inter-channel delay information (23), the K channels are encoded into the data stream (12).

80. The encoder (200) according to any one of claims 53 to 67. The multi-channel digital signals mentioned above include biological and physiological signals.

81. The encoder (200) according to any one of claims 53 to 68 is configured as follows: The data stream (12) includes channel permutation information (52) that signals are notified. The channel permutation information allows the channels of the multi-channel digital signal (10) to be permuted (50) in order to obtain the multi-channel digital signal (10).

82. The encoder (200) according to claim 69, wherein: The channel permutation information (52) describes the permutation of the channels in order to obtain a predetermined representation of the multi-channel digital signal (10).

83. The encoder (200) according to any one of claims 65 to 82 is configured as follows: Each of the Q sets is encoded into a substream (28) of the data stream associated with the corresponding set at time intervals (26), wherein: Each substream is formed by a sequence of substream portions (30), the sequence of which has consecutive time intervals encoded therein as associated with a set of corresponding substreams, and The substream portions of the substream are interleaved such that substream portions with overlapping time intervals are adjacent and continuous in the data stream, and that the substream portions with overlapping time intervals precede the substream portions with overlapping time intervals.

84. The encoder (200) according to claim 83, wherein: In each substream, some substream portions of the sequence of substream portions are encoded as random access points.

85. The encoder (200) according to claim 84, wherein: The random access points of the substreams are time-aligned.

86. The encoder (200) according to claim 84, wherein: The random access points of the substream are not aligned in time.

87. A decoder (100) for decoding a multi-channel digital signal (10) from a data stream (12), configured to: Decode the channel permutation information (52) from the data stream (12), and Decode N channels (20) of the multi-channel digital signal (10) from the data stream (12); and The K channels of the multi-channel digital signal (10) are permuted (50) according to the channel permutation information (52) in order to obtain the multi-channel digital signal (10).

88. The decoder (100) according to claim 87, wherein: The channel permutation information (52) describes the permutation of N channels in order to obtain a predetermined representation of the multi-channel digital signal (10).

89. The decoder (100) according to claim 87 or 88 is configured to decode N channels (20) of the multi-channel digital signal (10) from the data stream (12) using a multi-channel decoding scheme.

90. The decoder (100) according to claim 87 or 88 is configured to decode N channels (20) of the multi-channel digital signal (10) from the data stream (12) in the following manner: The K' encoded channels (14) representing the multi-channel digital signal (10) are grouped into Q sets, each set having n i There are n encoding channels, where n i Indicates the number of encoded channels in set i, where Q ≥ i ≥ 1; and A multi-channel decoding scheme is used to decode at least one set j of the Q sets from the data stream (12), where n j ≥2.

91. The decoder (100) according to claim 90 is configured to: The multi-channel decoding scheme is used to decode each of the Q sets j from the data stream (12), where n j ≥2.

92. The decoder (100) according to any one of claims 89 to 91, wherein: The multi-channel decoding scheme involves decoding m-channel downmixed signals from the data stream (12), where m < n. j The n-channel downmixed signal is obtained by upmixing the m-channel downmixed signal using the auxiliary information contained in the data stream (12). j Encoding channels, in order to obtain the n j One encoding channel.

93. The decoder (100) according to claim 92, wherein the auxiliary information includes one or more of the following: Inter-channel coherence ICC data, Channel level difference (CLD) data, and Channel prediction coefficients (CPC) data.

94. The decoder (100) according to any one of claims 89 to 93, wherein: The multi-channel decoding scheme is a multi-channel audio decoding scheme.

95. The decoder (100) according to any one of claims 89 to 94, wherein: The multi-channel decoding scheme involves: Transform-based decoding, comprising: deriving a scaling factor and transform coefficients from the data stream (12); using the scaling factor to perform spectral shaping on the transform coefficients to obtain a shaped spectrum; re-transforming the shaped spectrum to obtain a frame signal; and subjecting the frame signal to overlapping and addition processing; or Linear predictive coding (LPC) decoding, wherein the LPC decoding includes: deriving LPC coefficients and information about the residual signal from the data stream, and using the LPC coefficients to subject the residual signal to LPC synthesis.

96. The decoder (100) according to claim 95, wherein: Derivation of the scaling factor from the data stream (12) involves: entropy decoding of the scaling factor from the data stream (12), or decoding of LPC coefficients from the data stream (12) and converting the LPC coefficients into the scaling factor.

97. The decoder (100) according to any one of claims 94 to 96, wherein: The multi-channel decoding scheme is designed to handle a maximum of MAX encoded channels to the greatest extent possible, and The decoder is configured to group the K' encoded channels (14) representing the multi-channel digital signal (10) into Q sets, each set having n i n encoding channels, such that for each 0 < i < Q+1, n i ≤MAX.

98. The decoder (100) according to any one of claims 87 to 97 is configured to decode N channels (20) of the multi-channel digital signal (10) from the data stream (12) in such a way as follows: Decoding (29) from the data stream (12) to represent the K' encoded channels (14) of the multi-channel digital signal (10); and The K' encoded channels (14) representing the multi-channel digital signal (10) are subjected to channel retransformation (16), which retransforms the co-aligned sample positions (18) of the K' encoded channels (14) to obtain the K channels of the multi-channel digital signal (10).

99. The decoder (100) according to claim 98 is configured to derive the channel retransformation (16) from the transformation information (22) in the data stream (12).

100. The decoder (100) of claim 99 is configured to update the channel retransformation (16) based on the transformation information (22) in the data stream (12).

101. The decoder (100) according to any one of claims 98 to 100, wherein the channel retransformation (16) involves a sequence of partially channel retransformations.

102. The decoder (100) according to claim 101, wherein: The partial channel retransformation has different dimensions in the number of retransformed channels, so that different encoded channels are affected by different subsets of transformations, different numbers of transformations, or different subsequences of transformations in the sequence of partial channel retransformation.

103. The decoder (100) according to any one of claims 87 to 102, wherein the decoder is configured to: Decode the inter-channel delay information (23) from the data stream (12), and The K channels are delayed to each other according to the inter-channel delay information (23).

104. The decoder (100) according to any one of claims 87 to 103. The multi-channel digital signal (10) mentioned therein includes (for example) biological physiological signals.

105. The decoder (100) according to any one of claims 90 to 104 is configured to: Decode each of the Q sets from the substream (28) associated with the corresponding set of the data stream in units of time intervals (26), where: Each substream is formed by a sequence of substream portions (30), the sequence of which has consecutive time intervals encoded therein as associated with a set of corresponding substreams, and The substream portions of the substream are interleaved such that substream portions with overlapping time intervals are adjacent and continuous in the data stream, and that the substream portions with overlapping time intervals precede the substream portions with overlapping time intervals.

106. The decoder (100) according to claim 105, wherein: In each substream, some substream portions of the sequence of substream portions are encoded as random access points.

107. The decoder (100) according to claim 106, wherein: The random access points of the substreams are time-aligned.

108. The decoder (100) according to claim 106, wherein: The random access points of the substream are not aligned in time.

109. An encoder (200) for encoding a multi-channel digital signal (10) into a data stream (12), configured to: Encode the channel permutation information (52) into the data stream (12), and The N channels (20) of the multi-channel digital signal (10) are encoded into the data stream (12) in the following manner: according to the manner, the K channels of the multi-channel digital signal (10) are permuted according to the channel permutation information (52) in order to obtain the multi-channel digital signal (10).

110. The encoder (200) according to claim 109, wherein: The channel permutation information (52) describes the permutation between the coded N channels and the order of the N channels according to the predetermined representation of the multi-channel digital signal (10).

111. The encoder (200) according to claim 109 or 110 is configured to encode N channels (20) of the multi-channel digital signal (10) into the data stream (12) using a multi-channel encoding scheme.

112. The encoder (200) according to claim 109 or 110 is configured to encode N channels (20) of the multi-channel digital signal (10) into the data stream (12) in the following manner: The K' encoded channels (14) representing the multi-channel digital signal (10) are grouped into Q sets, each set having n i There are n encoding channels, where n i Indicates the number of encoded channels in set i, where Q ≥ i ≥ 1; and At least one set j of the Q sets is encoded into the data stream (12) using a multi-channel encoding scheme, wherein, n j ≥2。 113. The encoder (200) according to claim 112 is configured to: The multi-channel encoding scheme is used to encode each set j of the Q sets into the data stream (12), where n j ≥2.

114. The encoder (200) according to any one of claims 111 to 113, wherein: The multi-channel encoding scheme involves encoding m channels of downmixed signal into the data stream (12), where m < n. j By using the auxiliary information contained in the data stream (12) to upmix the m-channel downmixed signal, n can be obtained. j One encoding channel.

115. The encoder (200) according to claim 114, wherein the auxiliary information includes one or more of the following: Inter-channel coherence ICC data, Channel level difference (CLD) data, and Channel prediction coefficients (CPC) data.

116. The encoder (200) according to any one of claims 111 to 115, wherein: The multi-channel coding scheme is a multi-channel audio coding scheme.

117. The encoder (200) according to any one of claims 111 to 116. The multi-channel coding scheme mentioned above involves: Transform-based coding, wherein the transform-based coding includes: The scaling factor and transform coefficients are inserted into the data stream (12) such that the transform coefficients are spectrally shaped using the scaling factor to obtain a shaped spectrum, the shaped spectrum is re-transformed to obtain a frame signal, and the frame signal is subjected to overlapping addition processing to achieve reconstruction; or Linear predictive coding (LPC) includes inserting LPC coefficients and information about the residual signal into the data stream (12) such that the residual signal can be reconstructed by subjecting it to LPC synthesis using the LPC coefficients.

118. The encoder (200) according to claim 117, wherein: Inserting the scaling factor into the data stream (12) involves either encoding the scaling factor entropy into the data stream (12) or encoding the LPC coefficients into the data stream (12) so that the scaling factor can be obtained by converting the LPC coefficients into the scaling factor.

119. The encoder (200) according to any one of claims 112 to 118, wherein: The multi-channel encoding scheme is designed to handle a maximum of MAX encoding channels to the greatest extent possible, and The encoder is configured to group the K' encoded channels (14) representing the multi-channel digital signal (10) into Q sets, each set having n i n encoding channels, such that for each 0 < i < Q+1, n i ≤MAX.

120. The encoder (200) according to any one of claims 109 to 119 is configured to encode N channels (20) of the multi-channel digital signal (10) into the data stream (12) in such a way that: The K channels are subjected to channel transformation, which transforms the co-aligned sample positions of the K channels to obtain K' encoded channels (14) representing the multi-channel digital signal (10). The K' encoding channels (14) representing the multi-channel digital signal (10) are encoded into the data stream (12).

121. The encoder (200) according to claim 120 is configured to insert transformation information (22) into the data stream (12), from which a channel retransformation (16) corresponding to the channel transformation can be derived.

122. The encoder (200) according to claim 121 is configured to signal updates of the channel retransformation (16) in the data stream (12).

123. The encoder according to any one of claims 120 to 122, wherein the channel transformation involves a sequence of partial channel transformations (24).

124. The encoder (200) according to claim 123, wherein: The partial channel transformation (24) has different dimensions in the number of transformation channels, so that different coding channels are affected by different transformation subsets, different transformation numbers or different transformation subsequences in the sequence of the partial channel transformation (24).

125. The encoder (200) according to any one of claims 109 to 124 is configured to: The inter-channel delay information (23) is encoded into the data stream (12), and The K channels, which are mutually delayed according to the inter-channel delay information (23), are encoded into the data stream (12).

126. The encoder (200) according to any one of claims 109 to 125, wherein: The multi-channel digital signal (10) includes biological and physiological signals.

127. The encoder (200) according to any one of claims 112 to 126 is configured to: Each of the Q sets is encoded into a substream (28) of the data stream associated with the corresponding set at time intervals (26), wherein: Each substream is formed by a sequence of substream portions (30), the sequence of which has consecutive time intervals encoded therein as associated with a set of corresponding substreams, and The substream portions of the substream are interleaved such that substream portions with overlapping time intervals are adjacent and continuous in the data stream, and that the substream portions with overlapping time intervals precede the substream portions with overlapping time intervals.

128. The encoder (200) according to claim 127, wherein: In each substream, some substream portions of the sequence of substream portions are encoded as random access points.

129. The encoder (200) according to claim 128, wherein: The random access points of the substreams are time-aligned.

130. The encoder (200) according to claim 128, wherein: The random access points of the substream are not aligned in time.

131. A decoder (100) for decoding a multi-channel digital signal (10) from a data stream (12), configured to: Decode the inter-channel delay information (23) from the data stream (12), and Decode K channels of the multi-channel digital signal (10) from the data stream (12), and The K channels are delayed to each other according to the inter-channel delay information (23).

132. The decoder (100) according to claim 131 is configured to: The K channels are decoded using a multi-channel decoding scheme.

133. The decoder (100) according to claim 132. The multi-channel decoding scheme mentioned above involves: Transform-based multi-channel decoding, the transform-based multi-channel decoding includes: The scaling factor and transformation coefficient are derived from the data stream (12), and the transformation coefficient is spectrally shaped using the scaling factor to obtain a shaped spectrum. The shaped spectrum is then re-transformed to obtain a frame signal, and the frame signal is subjected to overlapping addition processing. or Linear predictive coding (LPC) decoding, wherein the LPC decoding includes: deriving LPC coefficients and information about the residual signal from the data stream (12), and using the LPC coefficients to subject the residual signal to LPC synthesis.

134. The decoder (100) according to claim 132, wherein: Derivation of the scaling factor from the data stream (12) involves: entropy decoding of the scaling factor from the data stream (12), or decoding of LPC coefficients from the data stream (12) and converting the LPC coefficients into the scaling factor.

135. The decoder (100) according to any one of claims 131 to 134. The multi-channel digital signal (10) mentioned therein includes biological and physiological signals.

136. The decoder (100) according to any one of claims 131 to 135. Decoding the K channels from the data stream (12) includes: A multi-channel decoding scheme (29) is used to decode n>1 encoded channels representing the multi-channel digital signal (10) from the data stream (12), the multi-channel decoding scheme involving: decoding (31) m-channel downmixed signal (32) from the data stream (12), where 0<m<n, and upmixing (33) the m-channel downmixed signal using auxiliary information (34) contained in the data stream (12) to obtain the n encoded channels.

137. The decoder (100) according to any one of claims 131 to 136. Decoding the K channels from the data stream (12) includes: The K' encoded channels (14) representing the multi-channel digital signal (10) are grouped into Q sets, each set having n i There are n encoding channels, where n i Indicates the number of encoded channels in set i, where Q ≥ i ≥ 1, and A multi-channel decoding scheme is used to decode at least one set j of the Q sets from the data stream (12), where n j ≥2.

138. The decoder (100) according to any one of claims 131 to 137. Decoding the K channels from the data stream (12) includes: The data stream (12) is decoded using a multi-channel decoding scheme to represent n≤K' encoded channels of the multi-channel digital signal (10).

139. The decoder (100) according to claim 137 or 138 is configured to: The K' encoded channels (14) of the multi-channel digital signal (10) are subjected to channel retransformation (16), which retransforms the co-aligned sample positions of the K' encoded channels (14) to obtain the K channels of the multi-channel digital signal (10).

140. The decoder (100) of claim 139 is configured to derive the channel retransformation (16) from the transformation information (22) in the data stream (12).

141. The decoder (100) of claim 140 is configured to update the channel retransformation (16) based on the transformation information (22) in the data stream (12).

142. The decoder (100) according to any one of claims 139 to 141, wherein the channel retransformation (16) involves a sequence of partially channel retransformations.

143. The decoder (100) according to claim 142, wherein: The partial channel retransformation has different dimensions in the number of retransformed channels, so that different encoded channels are affected by different subsets of transformations, different numbers of transformations, or different subsequences of transformations in the sequence of partial channel retransformation.

144. The decoder (100) according to any one of claims 137 to 143, wherein: The multi-channel decoding scheme is designed to handle a maximum of MAX encoded channels to the greatest extent possible. The decoder is configured to group the K' encoded channels (14) representing the multi-channel digital signal (10) into Q sets, each set having n i n encoding channels, such that for each 0 < i < Q+1, n i ≤MAX.

145. The decoder (100) according to any one of claims 137 to 144 is configured to: Decode channel grouping information from the data stream (12) and perform grouping using the channel grouping information.

146. The decoder (100) according to claim 145, wherein: The channel grouping information includes: Indicates the number n of the encoded channels in the Q sets. i syntax elements, and / or The sequential channel order among the K' encoded channels (14) is used to group the K' encoded channels (14) into the Q sets such that the encoded channels of different sets in the Q sets are not interleaved along the sequential channel order, and one or more encoded channels in a set are adjacent and continuous in the sequential channel order.

147. The decoder (100) according to any one of claims 136 to 146, wherein the auxiliary information includes one or more of the following: Inter-channel coherence ICC data, Channel level difference (CLD) data, and Channel prediction coefficients (CPC) data.

148. The decoder (100) according to any one of claims 131 to 147 is configured to: The K channels of the multi-channel digital signal (10) are permuted (50) according to the channel permutation information (52) notified in the data stream (12) in order to obtain the multi-channel digital signal (10).

149. The decoder (100) according to claim 148, wherein: The channel permutation information (52) describes the permutations of the N channels obtained by the channel retransformation (16) in order to obtain a predetermined representation of the multi-channel digital signal (10).

150. The decoder (100) according to any one of claims 137 to 149 is configured to: Decode each of the Q sets from the substream (28) associated with the corresponding set of the data stream in units of time intervals (26), where: Each substream is formed by a sequence of substream portions (30), the sequence of which has consecutive time intervals encoded therein as associated with a set of corresponding substreams, and The substream portions of the substream are interleaved such that substream portions with overlapping time intervals are adjacent and continuous in the data stream, and that the substream portions with overlapping time intervals precede the substream portions with overlapping time intervals.

151. The decoder (100) according to claim 150, wherein: In each substream, some substream portions of the sequence of substream portions are encoded as random access points.

152. The decoder (100) according to claim 151, wherein: The random access points of the substreams are time-aligned.

153. The decoder (100) according to claim 151, wherein: The random access points of the substream are not aligned in time.

154. An encoder (200) for encoding multi-channel digital signals (10) into a data stream, configured to: The inter-channel delay information (23) is encoded into the data stream (12), and With the K channels of the multi-channel digital signal (10) mutually delayed according to the inter-channel delay information (23), the K channels are encoded into the data stream (12).

155. The encoder (200) according to claim 154 is configured to: The K channels are encoded using a multi-channel encoding scheme.

156. The encoder (200) according to claim 155. The multi-channel coding scheme mentioned above involves: Transform-based multichannel coding, the transform-based multichannel coding includes: The scaling factor and transform coefficients are inserted into the data stream (12) such that the transform coefficients are spectrally shaped using the scaling factor to obtain a shaped spectrum, the shaped spectrum is re-transformed to obtain a frame signal, and the frame signal is subjected to overlapping addition processing to achieve reconstruction; or Linear predictive coding (LPC) encoding includes encoding LPC coefficients and information about the residual signal into the data stream, such that decoding can be achieved by subjecting the residual signal to LPC synthesis using the LPC coefficients.

157. The encoder (200) according to claim 156, wherein: Inserting the scaling factor into the data stream (12) involves either encoding the scaling factor entropy into the data stream (12) or encoding the LPC coefficients into the data stream (12) so that the scaling factor can be derived by converting the LPC coefficients into the scaling factor.

158. The encoder (200) according to any one of claims 154 to 157. The multi-channel digital signal (10) mentioned therein includes biological and physiological signals.

159. The encoder (200) according to any one of claims 154 to 158. Encoding the K channels into the data stream (12) includes: A multi-channel encoding scheme is used to encode n>1 of the K' encoding channels (14) representing the multi-channel digital signal (10) into the data stream (12), where n≤K'. The multi-channel encoding scheme involves encoding auxiliary information and an m-channel downmixed signal (32) into the data stream (12), where 0<m<n, such that the n encoding channels can be decoded by upmixing (33) the m-channel downmixed signal using the auxiliary information (34) contained in the data stream (12).

160. The encoder (200) according to any one of claims 154 to 159. Encoding the K channels into the data stream (12) includes: The K' encoded channels (14) representing the multi-channel digital signal (10) are grouped into Q sets, each set having n i There are n encoding channels, where n i Indicates the number of encoded channels in set i, where Q ≥ i ≥ 1, and At least one set j from the Q sets is encoded into the data stream (12) using a multi-channel encoding scheme, where n j ≥2.

161. The encoder (200) according to any one of claims 154 to 160. Encoding the K channels into the data stream (12) includes: The data stream (12) is encoded from the data stream (12) using n≤K' encoded channels to represent the multi-channel digital signal (10).

162. The encoder (200) according to claim 160 or 160 is configured to: The K channels are subjected to channel transformation, which transforms the co-aligned sample positions of the K channels to obtain K' encoded channels (14) representing the multi-channel digital signal (10).

163. The encoder (200) according to claim 162 is configured to insert transformation information (22) into the data stream (12), from which a channel retransformation (16) corresponding to the channel transformation can be derived.

164. The encoder (200) according to claim 163 is configured to change the channel transformation and send an update of the channel retransformation (16) in the data stream (12) to reflect the change.

165. The encoder (200) according to any one of claims 162 to 164, wherein the channel transformation involves a sequence of partial channel transformations (24).

166. The encoder (200) according to claim 165, wherein: The partial channel transformation (24) has different dimensions in the number of transformation channels, so that different coding channels are affected by different transformation subsets, different transformation numbers, or different transformation subsequences in the sequence of partial channel retransformation.

167. The encoder (200) according to claim 160, wherein: The multi-channel coding scheme is designed to handle a maximum of MAX coding channels to the greatest extent possible. The encoder is configured to group the K' encoded channels (14) representing the multi-channel digital signal (10) into Q sets, each set having n i n encoding channels, such that for each 0 < i < Q+1, n i ≤MAX.

168. The encoder (200) according to any one of claims 160 to 167 is configured to: The channel grouping information is encoded into the data stream (12), and grouping is performed according to the channel grouping information.

169. The encoder (200) according to claim 168, wherein: The channel grouping information includes: Indicates the number n of the encoded channels in the Q sets. i syntax elements, and / or The sequential channel order among the K' encoded channels (14) is used to group the K' encoded channels (14) into the Q sets such that the encoded channels of different sets in the Q sets are not interleaved along the sequential channel order, and one or more encoded channels in a set are adjacent to each other in the sequential channel order.

170. The encoder (200) according to any one of claims 159 to 169, wherein the auxiliary information includes: Inter-channel coherence ICC data, Channel level difference (CLD) data, and Channel prediction coefficients (CPC) data.

171. The encoder (200) according to any one of claims 154 to 170 is configured to: In the data stream (12), a channel replacement information (52) is signaled. According to the channel replacement information, K channels of the multi-channel digital signal (10) will be replaced so that the multi-channel digital signal (10) can be obtained from the data stream (12).

172. The encoder (200) according to claim 171, wherein: The channel permutation information (52) describes the permutations of the N channels obtained by the channel retransformation (16) in order to obtain a predetermined representation of the multi-channel digital signal (10).

173. The encoder (200) according to any one of claims 160 to 172 is configured to: Each of the Q sets is encoded into a substream (28) of the data stream associated with the corresponding set at time intervals (26), wherein: Each substream is formed by a sequence of substream portions (30), the sequence of which has consecutive time intervals encoded therein as associated with a set of corresponding substreams, and The substream portions of the substream are interleaved such that substream portions with overlapping time intervals are adjacent and continuous in the data stream, and that the substream portions with overlapping time intervals precede the substream portions with overlapping time intervals.

174. The encoder (200) according to claim 173, wherein: In each substream, some substream portions of the sequence of substream portions are encoded as random access points.

175. The encoder (200) according to claim 174, wherein: The random access points of the substreams are time-aligned.

176. The encoder (200) according to claim 174, wherein: The random access points of the substream are not aligned in time.

177. A decoder (100) for decoding a multi-channel digital signal (10) from a data stream (12), configured to: Decoding (29) from the data stream (12) to represent the K' encoded channels (14) of the multi-channel digital signal (10); and The K' encoded channels (14) are subjected to channel retransformation (16), which retransforms the co-aligned sample positions (18) of the K' encoded channels (14) to obtain K channels (20) of the multi-channel digital signal (10).

178. The decoder (100) according to claim 177, wherein K = K'.

179. The decoder (100) according to claim 177 or 178, wherein the channel retransformation (16) is linear.

180. The decoder (100) according to any one of claims 17 to 179 is configured to derive the channel retransformation (16) from the transformation information (22) in the data stream (12).

181. The decoder (100) according to claim 180 is configured to update the channel retransformation (16) based on the transformation information (22) in the data stream (12).

182. The decoder (100) according to any one of claims 177 to 181 is configured to: The K' encoded channels (14) of the multi-channel digital signal (10) are decoded from the data stream (12) using an audio decoding scheme.

183. The decoder (100) according to claim 182, wherein: The audio decoding scheme is a multi-channel audio decoding scheme.

184. The decoder (100) according to claim 183 is configured to: Decoding schemes are used to decode K' encoded channels (14) of the multi-channel digital signal (10) from the data stream (12), wherein the decoding schemes involve: Transform-based decoding, which includes: The scaling factor and transformation coefficient are derived from the data stream (12), and the transformation coefficient is spectrally shaped using the scaling factor to obtain a shaped spectrum. The shaped spectrum is then re-transformed to obtain a frame signal, and the frame signal is subjected to overlapping addition processing. or Linear predictive coding (LPC) decoding, wherein the LPC decoding includes: deriving LPC coefficients and information about the residual signal from the data stream, and using the LPC coefficients to subject the residual signal to LPC synthesis.

185. The decoder (100) according to claim 184, wherein: Derivation of the scaling factor from the data stream (12) involves: entropy decoding of the scaling factor from the data stream (12), or decoding of LPC coefficients from the data stream (12) and converting the LPC coefficients into the scaling factor.

186. The decoder (100) according to any one of claims 177 to 185 is configured to: A multi-channel decoding scheme is used to decode K' encoded channels (14) of the multi-channel digital signal (10) from the data stream (12), the multi-channel decoding scheme comprising: Decode (31) the m-channel downmixed signal (32) from the data stream (12), where 0 < m < n, and upmix (33) the m-channel downmixed signal by using the auxiliary information (34) contained in the data stream (12) to obtain n encoded channels, where n ≤ K'.

187. The decoder (100) according to any one of claims 177 to 186 is configured to: The K' encoded channels (14) of the multi-channel digital signal (10) are decoded from the data stream (12) in the following manner: The K' encoded channels (14) are grouped (21) into Q sets (25), each set having n i There are n encoding channels, where n i Indicates the number of encoded channels in set i, where Q ≥ i ≥ 1, and A multi-channel decoding scheme is used to decode at least one set j of the Q sets from the data stream (12), where n j ≥2, the multi-channel decoding scheme involves: decoding m-channel downmixed signals (32) from the data stream (12), where m < n j The set j is obtained by upmixing the m-channel downmixed signal using the auxiliary information (34) contained in the data stream (12). j One encoding channel (14).

188. The decoder (100) according to claim 187, wherein: The multi-channel decoding scheme is designed to handle a maximum of MAX encoded channels to the greatest extent possible. The decoder is configured to group the K' encoded channels (14) into Q sets, each set having n i n encoding channels, such that for each 0 < i < Q+1, n i ≤MAX.

189. The decoder (100) according to any one of claims 186 to 188, wherein the auxiliary information includes one or more of the following: Inter-channel coherence ICC data, Channel level difference (CLD) data, and Channel prediction coefficients (CPC) data.

190. The decoder (100) according to any one of claims 187 to 189 is configured to: Decode channel grouping information (40) from the data stream (12) and perform grouping using the channel grouping information.

191. The decoder (100) according to claim 190, wherein, The channel grouping information (40) includes: Indicates the number n of the encoded channels in the Q sets. i Syntax elements; and / or The sequential channel order among the K' encoding channels (14) is used to group the K' encoding channels (14) into the Q sets, such that the encoding channels of different sets in the Q sets are not interleaved along the sequential channel order, and one or more encoding channels in a set are adjacent and continuous in the sequential channel order.

192. The decoder (100) according to any one of claims 177 to 191 is configured to: The K channels of the multi-channel digital signal (10) are permuted (50) according to the channel permutation information (52) notified in the data stream (12) in order to obtain the multi-channel digital signal (10).

193. The decoder (100) according to claim 192, wherein: The channel permutation information (52) describes the permutations of the N channels obtained by the channel retransformation (16) in order to obtain a predetermined representation of the multi-channel digital signal (10).

194. The decoder (100) according to any one of claims 177 to 193. The multi-channel digital signal (10) mentioned therein includes biological and physiological signals.

195. The decoder (100) according to any one of claims 177 to 194, wherein: The channel retransformation involves a sequence of partial channel retransformations.

196. The decoder (100) according to claim 195, wherein: The partial channel retransformation has different dimensions in the number of retransformed channels, so that different encoded channels are affected by different subsets of transformations, different numbers of transformations, or different subsequences of transformations in the sequence of partial channel retransformation.

197. The decoder (100) according to any one of claims 177 to 196, wherein the decoder is configured to: Decode the inter-channel delay information from the data stream; and delay the K channels to each other based on the inter-channel delay information.

198. The decoder (100) according to claim 197, wherein the decoder is configured to: Based on the inter-channel delay information, the K channels are mutually delayed in such a way that the mutual delay between two channels is greater than a sample interval between the samples of the two channels.

199. The decoder (100) according to any one of claims 187 to 198 is configured to: Decode each of the Q sets from the substream (28) associated with the corresponding set of the data stream in units of time intervals (26), where: Each substream is formed by a sequence of substream portions (30), the sequence of which has consecutive time intervals encoded therein as associated with a set of corresponding substreams, and The substream portions of the substream are interleaved such that substream portions with overlapping time intervals are adjacent and continuous in the data stream, and that the substream portions with overlapping time intervals precede the substream portions with overlapping time intervals.

200. The decoder (100) according to claim 199, wherein: In each substream, some substream portions of the sequence of substream portions are encoded as random access points.

201. The decoder (100) according to claim 200, wherein: The random access points of the substreams are time-aligned.

202. The decoder (100) according to claim 200, wherein: The random access points of the substream are not aligned in time.

203. An encoder (200) for encoding a multi-channel digital signal (10) into a data stream (12), configured to: The K channels (20) of the multi-channel digital signal (10) undergo channel transformation, which transforms the co-aligned sample positions of the K channels to obtain K' encoded channels (14). The K' encoded channels (14) representing the multi-channel digital signal (10) are encoded into the data stream (12).

204. The encoder (200) according to claim 203, wherein K = K'.

205. The encoder (200) according to claim 203 or 204, wherein the channel transformation is linear.

206. The encoder (200) according to any one of claims 203 to 205 is configured to insert transformation information (22) into the data stream (12), from which a channel retransformation (16) corresponding to the channel transformation can be derived.

207. The encoder (200) according to claim 206 is configured to change the channel transformation and signal the update of the channel re-transformation (16).

208. The encoder (200) according to any one of claims 203 to 207 is configured to: The K' encoded channels (14) of the multi-channel digital signal (10) are encoded into the data stream (12) using an audio encoding scheme.

209. The encoder (200) according to claim 208, wherein: The audio encoding scheme is a multi-channel audio encoding scheme.

210. The encoder (200) according to any one of claims 203 to 209 is configured to: One or more channels of the K channels of the multi-channel digital signal (10) are resampled so that: The K channels, or one or more sets of the K channels, are sampled from each other at equal frequencies and equal sampling phases; and / or The K channels, or one or more sets of the K channels, are sampled from each other at equal frequencies that are integer multiples of the common fundamental frequency and their phases are adjusted.

211. The encoder (200) according to claim 209 or 210 is configured to: The K' encoded channels (14) of the multi-channel digital signal (10) are encoded into the data stream (12) using an encoding scheme, the encoding scheme involving: Transform-based coding, wherein the transform-based coding includes: The scaling factor and transformation coefficient are inserted into the data stream (12), and the transformation coefficient is spectrally shaped by the scaling factor to obtain a shaped spectrum. The shaped spectrum is then re-transformed to obtain a frame signal, and the frame signal is subjected to overlapping addition processing so that the reconstruction can be obtained from the scaling factor and transformation coefficient. or Linear predictive coding (LPC) includes inserting LPC coefficients and information about the residual signal into the data stream (12) such that the residual signal can be reconstructed by subjecting it to LPC synthesis using the LPC coefficients.

212. The encoder (200) according to claim 211, wherein: Inserting the scaling factor into the data stream (12) involves: encoding the scaling factor entropy into the data stream (12), or encoding the LPC coefficients into the data stream (12), by converting the LPC coefficients into the scaling factor to obtain the scaling factor.

213. The encoder (200) according to any one of claims 203 to 212 is configured to: The K' encoding channels (14) of the multi-channel digital signal (10) are encoded into the data stream (12) using a multi-channel encoding scheme, wherein the multi-channel encoding scheme includes: The auxiliary information and the m-channel downmixed signal (32) are encoded into the data stream (12), where 0 < m < n. The m-channel downmixed signal is upmixed (33) using the auxiliary information (34) contained in the data stream (12) so that n encoded channels can be obtained, where n ≤ K'.

214. The encoder (200) according to any one of claims 203 to 213 is configured to: The K' encoded channels (14) of the multi-channel digital signal (10) are encoded into the data stream (12) in the following manner: The K' encoded channels (14) are grouped (21) into Q sets, each set having n i There are n encoding channels, where n i Indicates the number of encoded channels in set i, where Q ≥ i ≥ 1, and At least one set j from the Q sets is encoded into the data stream (12) using a multi-channel encoding scheme, where n j ≥2, the multi-channel encoding scheme involves encoding auxiliary information and m-channel downmixed signals (32) into the data stream (12), where m < n j By using the auxiliary information (34) contained in the data stream (12) to upmix the m-channel downmixed signal, the n-value of set j can be obtained. j One encoding channel (14).

215. The encoder (200) according to claim 214, wherein The multi-channel encoding scheme is designed to handle a maximum of MAX encoding channels to the greatest extent possible, and The encoder is configured to group the K' encoded channels (14) into Q sets, each set having n i n encoding channels, such that for each 0 < i < Q+1, n i ≤MAX.

216. The encoder (200) according to any one of claims 213 to 215, wherein the auxiliary information includes one or more of the following: Inter-channel coherence ICC data, Channel level difference (CLD) data, and Channel prediction coefficients (CPC) data.

217. The encoder (200) according to any one of claims 214 to 216 is configured to: The channel grouping information (40) is encoded into the data stream (12), and grouping is performed according to the channel grouping information.

218. The encoder (200) according to claim 217, wherein: The channel grouping information (40) includes: Indicates the number n of the encoded channels in the Q sets. i Syntax elements; and / or The sequential channel order among the K' encoding channels (14) is used to group the K' encoding channels (14) into the Q sets, such that the encoding channels of different sets in the Q sets are not interleaved along the sequential channel order, and one or more encoding channels in a set are adjacent and continuous in the sequential channel order.

219. The encoder (200) according to any one of claims 203 to 218 is configured to: The K channels of the multi-channel digital signal (10) are permuted (50), and the permutation is notified using the channel permutation information (52) signal in the data stream (12).

220. The encoder (200) according to claim 219, wherein: The channel permutation information (52) describes the permutations of the N channels encoded into the data stream (12) to obtain a predetermined representation of the multi-channel digital signal (10).

221. The encoder (200) according to any one of claims 203 to 220. The multi-channel digital signal (10) mentioned therein includes biological and physiological signals.

222. The encoder (200) according to any one of claims 203 to 221, wherein: The channel transformation involves a sequence of partial channel transformations (24).

223. The encoder (200) according to claim 222, wherein: The partial channel transformation (24) has different dimensions in the number of transformation channels, so that different coding channels are affected by different transformation subsets, different transformation numbers or different transformation subsequences in the sequence of the partial channel transformation (24).

224. The encoder (200) according to any one of claims 203 to 223, wherein the multi-channel digital signal has N channels, and the encoder is configured to: The inter-channel delay information is encoded into the data stream (12); and The N channels, which are mutually delayed according to the inter-channel delay information, are encoded into the data stream (12).

225. The encoder (200) according to any one of claims 224, wherein the multi-channel digital signal (10) has N channels, and the encoder is configured to: The N channels, which are mutually delayed according to the inter-channel delay information, are encoded into the data stream (12) in such a way that the mutual delay experienced by two of the K channels is greater than a sample interval between the samples of the two channels.

226. The encoder (200) according to any one of claims 214 to 225 is configured to: Each of the Q sets is encoded into a substream (28) of the data stream (12) associated with the corresponding set at time intervals (26), wherein: Each substream is formed by a sequence of substream portions (30), the sequence of which has consecutive time intervals encoded therein as associated with a set of corresponding substreams, and The substream portions of the substream are interleaved such that substream portions with overlapping time intervals are adjacent and continuous in the data stream, and that the substream portions with overlapping time intervals precede the substream portions with overlapping time intervals.

227. The encoder (200) according to claim 226, wherein: In each substream, some substream portions of the sequence of substream portions are encoded as random access points.

228. The encoder (200) according to claim 227, wherein: The random access points of the substreams are time-aligned.

229. The encoder (200) according to claim 228, wherein: The random access points of the substream are not aligned in time.

230. A method for decoding biological physiological signals (102) from a data stream (104; 12), the method comprising: Using an audio decoding scheme (106) from the data stream (104); 12) Decode the biological physiological signal (102).

231. A method for decoding a multi-channel digital signal (10) from a data stream (12), the method comprising: Decode the channel packet information (40) from the data stream (12); According to the channel grouping information (40), the K' encoded channels (14) representing the multi-channel digital signal (10) are grouped (21) into Q sets, each set having n i There are n encoding channels, where n i Indicates the number of encoded channels in set i, where Q ≥ i ≥ 1, and A multi-channel decoding scheme is used to decode at least one set j of the Q sets from the data stream (12), where n j ≥2.

232. A method for decoding a multi-channel digital signal (10) from a data stream (12), the method comprising: Decode the channel permutation information (52) from the data stream (12); and Decode N channels (20) of the multi-channel digital signal (10) from the data stream (12); and The K channels of the multi-channel digital signal (10) are permuted (50) according to the channel permutation information (52) in order to obtain the multi-channel digital signal (10).

233. A method for decoding a multi-channel digital signal (10) from a data stream (12), the method comprising: Decode the inter-channel delay information (23) from the data stream (12); and Decode the K channels of the multi-channel digital signal (10) from the data stream (12); and The K channels are delayed to each other according to the inter-channel delay information (23).

234. A method for decoding a multi-channel digital signal (10) from a data stream (12), the method comprising: Decoding (29) from the data stream (12) to represent the K' encoded channels (14) of the multi-channel digital signal (10); and The K' encoded channels (14) are subjected to channel retransformation (16), which retransforms the co-aligned sample positions (18) of the K' encoded channels (14) to obtain K channels (20) of the multi-channel digital signal (10).

235. A method for encoding biological physiological signals (102) into a data stream (104), the method comprising: The biophysiological signals are encoded into the data stream using an audio coding scheme (206).

236. A method for encoding a multi-channel digital signal (10) into a data stream (12), the method comprising: The channel grouping information is encoded into the data stream (12). Using the channel grouping information, it is indicated that the K' encoded channels (14) of the multi-channel digital signal (10) are grouped into Q sets, each set having n i There are n encoding channels, where n i Indicates the number of encoded channels in set i, where Q ≥ i ≥ 1, and At least one set j from the Q sets is encoded into the data stream (12) using a multi-channel encoding scheme, where n j ≥2.

237. A method for encoding a multi-channel digital signal (10) into a data stream (12), the method comprising: Encode the channel permutation information (52) into the data stream (12); and The N channels (20) of the multi-channel digital signal (10) are encoded into the data stream (12) in the following manner: according to the manner, the K channels of the multi-channel digital signal (10) are permuted according to the channel permutation information (52) in order to obtain the multi-channel digital signal (10).

238. A method for encoding a multi-channel digital signal (10) into a data stream, the method comprising: The inter-channel delay information (23) is encoded into the data stream (12); and The K channels of the multi-channel digital signal (10) are encoded into the data stream (12) under the condition that the K channels are mutually delayed according to the inter-channel delay information (23).

239. A method for encoding a multi-channel digital signal (10) into a data stream (12), the method comprising: The K channels (20) of the multi-channel digital signal (10) undergo channel transformation, which transforms the co-aligned sample positions of the K channels to obtain K' encoded channels (14). The K' encoding channels (14) representing the multi-channel digital signal (10) are encoded into the data stream (12).

240. A data stream (12) generated using the encoding method according to claim 235.

241. A data stream (12) generated using the encoding method according to claim 236.

242. A data stream (12) generated using the encoding method according to claim 237.

243. A data stream (12) generated using the encoding method according to claim 238.

244. A data stream (12) generated using the encoding method according to claim 239.

Images