28 results about "Linear prediction coefficient" patented technology

The invention belongs to the technical field of voice signal processing, and discloses a method and system for driving a virtual human by a voice signal, and the method comprises the steps: carrying out the resampling processing of original audio signal data, obtaining a voice signal, solving a linear prediction coefficient based on a dynamic framing strategy and a linear prediction analysis technology, and obtaining a linear prediction result; obtaining a voice feature sequence of the voice signal; performing enhancement processing on the voice feature sequence by using a dynamic feature enhancement technology, obtaining an enhanced feature sequence based on a processing result, and constructing an animation sequence prediction model by taking the enhanced feature sequence as input; and obtaining animation sequence parameters based on an output result of the animation sequence prediction model, and combining the animation sequence parameters with the virtual human role model to perform facial action synthesis and rendering operation. According to the invention, the earlier stage processing flow is greatly simplified, and the usability and applicability in the driving process are improved.

Embodiments according to the invention comprise an audio decoder configured to, for a predetermined frame among consecutive frames, decode, from a data stream, a quantized spectrum, a linear prediction coefficient based spectral envelope representation and a fundamental frequency related parameter. The decoder is configured to determine a spectral shaping function from the linear prediction coefficient based spectral envelope representation using a first manner below a pitch frequency determined from the fundamental frequency related parameter, and a second manner above the pitch frequency, to spectrally shape the quantized spectrum using the spectral shaping function to obtain a dequantized spectrum and to reconstruct the predetermined frame using the dequantized spectrum. Furthermore, the audio decoder is configured so that the spectral shaping function is, at a predetermined spectral position, lower if the pitch frequency is spectrally higher than the predetermined spectral position, than compared to if the pitch frequency is spectrally lower than the predetermined spectral position. Further decoders, encoders and methods are disclosed.

Provided are an apparatus for encoding an audio signal and a method of an operation thereof. An audio signal encoding method includes obtaining quantized linear prediction (LP) coefficients by performing a linear predictive coding (LPC) analysis and quantization on an input audio signal, generating a reference signal by applying discrete Fourier transform (DFT) to the input audio signal, obtaining LP residual coefficients from the reference signal, scaling magnitudes of the LP residual coefficients using the quantized LP coefficients and the reference signal, and quantizing phases of the LP residual coefficients and the scaled magnitudes of the LP residual coefficients.

The present invention belongs to the field of speech signal processing technology, and discloses a method and system for driving a virtual human with a speech signal, the method comprising: resampling the original audio signal data to obtain a speech signal, and solving the linear prediction coefficient based on a dynamic frame division strategy and a linear prediction analysis technology to obtain a speech feature sequence of the speech signal; using a dynamic feature enhancement technology to enhance the speech feature sequence, and obtaining an enhanced feature sequence based on the processing result, and using the enhanced feature sequence as input to construct an animation sequence prediction model; obtaining animation sequence parameters based on the output result of the animation sequence prediction model, and combining the animation sequence parameters with a virtual human role model to perform facial action synthesis and rendering operations. The present invention greatly simplifies the pre-processing process and improves the ease of use and applicability in the driving process.

Systems and methods for quantum linear prediction include autocorrelations formed with QFTs, and a modified quantum HHL circuit that includes appropriate normalization and encoding steps for solving a linear system of equations, including normalization of the quantum autocorrelation sequence using a norm factor; measuring a probabilistic distribution associated with values of a quantum state solution vector representing a set of quantum autoregressive parameters that correlate with a linear relationship between the quantum autocorrelation matrix and the quantum autocorrelation sequence; and generating a set of quantum linear prediction coefficients by re-normalization of the quantum state solution vector using the norm factor associated with the quantum autocorrelation sequence of the preprocessed input.

The invention discloses a broadband signal measurement method based on improved Prony, and the method comprises the following steps: 1, carrying out the singular value decomposition of a broadband signal, determining an initial singular value demarcation point through an energy accumulation method, and introducing a Pearson correlation coefficient to distinguish the main frequency components and noise of the signal, thereby determining the model order of Prony; 2, preprocessing the broadband signal by using variational mode decomposition, solving a variational problem through iteration, decomposing the complex broadband signal into a plurality of intrinsic mode functions with narrowband frequency spectrums, and reducing the complexity of signal processing; step 3, introducing a disturbance matrix satisfying a Toeplitz structure, iteratively optimizing a linear prediction coefficient of a Prony algorithm, using the Prony algorithm after parameter optimization to identify each mode, obtaining frequency, amplitude and phase information of each mode, and realizing measurement of broadband signals; according to the method, the problem of insufficient measurement precision caused by noise interference and difficulty in model order selection in broadband signal measurement of a traditional Prony algorithm is solved.

Embodiments according to the invention comprise an audio decoder configured to, for a predetermined frame among consecutive frames, decode, from a data stream, a quantized spectrum and a linear prediction coefficient based envelope representation. Furthermore, the decoder is configured to locate, in the quantized spectrum, zero-quantized portions and non-zero-quantized portions and to derive a dequantized spectrum using in zero-quantized portions of the quantized spectrum, filling the quantized spectrum with a synthesized spectral data modified depending, according to a first manner, on the linear prediction coefficient based envelope representation, and in non-zero-quantized portions of the quantized spectrum, modifying the quantized spectrum depending, in a second manner, on the linear prediction coefficient based envelope representation. In addition, the decoder is configured to reconstruct the predetermined frame using the dequantized spectrum. The audio decoder is configured so that, for a predetermined portion, the modification according to the first manner and the modification according to the second manner cause a spectral quantization noise shaping which comprises different smoothness characteristics. Beyond that, corresponding encoders and methods are disclosed.

To be able to appropriately change a dimension of a line spectrum pair.SOLUTION: A processing unit using line spectral pairs comprises: an LSP inverse transform section 27 (27-m, 27-m+1, ...) that transforms a first line spectral pair in a first dimension into a first linear prediction coefficient in the first dimension; an LSP dimensional transform section 28 (28-m, 28-m+1, ...) that transforms the first linear prediction coefficient in the first dimension into a second linear prediction coefficient in a second dimension; and an LSP dimensional transformation section 29 (29-m, 29-m+1, ...) that transforms the second linear prediction coefficient in the second dimension into a second line spectrum pair in the second dimension.SELECTED DRAWING: Figure 2

Embodiments of the present invention provide a linear prediction-based noise signal processing method, a linear prediction-based noise signal generation method, an encoder, a decoder, and an encoding and decoding system. The noise signal processing method according to the embodiments of the present invention includes: acquiring a noise signal, and obtaining a linear prediction coefficient according to the noise signal; filtering the noise signal according to the linear prediction coefficient, to obtain a linear prediction residual signal; obtaining a spectral envelope of the linear prediction residual signal according to the linear prediction residual signal; and encoding the spectral envelope of the linear prediction residual signal. According to the noise processing method, the noise generation method, the encoder, the decoder, and the encoding and decoding system that are in the embodiments of the present invention, more spectral details of an original background noise signal can be recovered, so that comfort noise can be closer to original background noise in terms of subjective auditory perception of a user, and subjective perception quality of the user is improved.

The present application provides a snoring detection method, apparatus, device and readable storage medium. The method includes: calculating the Mel-frequency cepstral coefficients and linear prediction coefficients of the microphone sound signal respectively; performing feature splicing and dimensionality reduction processing on the Mel-frequency cepstral coefficients and linear prediction coefficients to obtain a target feature sequence; calculating a target feature vector corresponding to the micro-Doppler feature of the target body part of the user based on the radar echo signal; wherein, the microphone sound signal and the radar echo signal are signals synchronously collected in the user's sleep environment; the micro-Doppler feature is used to reflect the vital activity state of the target part; inputting the target feature sequence and the target feature vector into a preset fusion processing model respectively to output a snoring detection result. Through the implementation of the solution of the present application, by fusing the feature data collected by two sensors, the snoring recognition ability in the home scenario can be effectively improved, and the complexity of data processing and the dependence of data processing on the network can be reduced.

The application belongs to the field of underwater target direction estimation, and combines a linear prediction algorithm with an ETAM algorithm of a vector hydrophone array to estimate the direction of an underwater target. Specifically, the application comprises the following steps: 1) constructing a data matrix of a virtual array after synthetic aperture by using the principle of the ETAM algorithm of the vector hydrophone array; 2) performing high-order linear prediction to solve forward and backward linear prediction coefficient vectors of the matrix; 3) extending the virtual array by k elements to the right or to the left; and 4) performing direction estimation on a signal matrix received by the extended virtual array in the previous step by using a conventional beam forming algorithm of the vector hydrophone array to obtain the target angle. The DOA estimation method of the moving platform vector hydrophone array based on the linear prediction and the ETAM algorithm can realize target direction estimation under the condition of lower signal-to-noise ratio and fewer array elements, and improves the target direction estimation precision and resolution of the underwater moving array.

A method and apparatus for audio signal encoding are provided. According to one embodiment, an audio signal decoding method includes an operation for receiving a bitstream including information related to a first audio signal. The method includes operations for generating a second audio signal based on a first spectrum obtained from the bitstream and first quantization information. The first quantization information includes quantization information generated based on a first complex polynomial having a complex linear prediction coefficient (LPC) corresponding to the first audio signal as a coefficient.

Embodiments of the present invention provide a linear prediction-based noise signal processing method, a linear prediction-based noise signal generation method, an encoder, a decoder, and an encoding and decoding system. The noise signal processing method according to the embodiments of the present invention includes: acquiring a noise signal, and obtaining a linear prediction coefficient according to the noise signal; filtering the noise signal according to the linear prediction coefficient, to obtain a linear prediction residual signal; obtaining a spectral envelope of the linear prediction residual signal according to the linear prediction residual signal; and encoding the spectral envelope of the linear prediction residual signal. According to the noise processing method, the noise generation method, the encoder, the decoder, and the encoding and decoding system that are in the embodiments of the present invention, more spectral details of an original background noise signal can be recovered, so that comfort noise can be closer to original background noise in terms of subjective auditory perception of a user, and subjective perception quality of the user is improved.

Embodiments according to the invention comprise an audio decoder configured to, for a predetermined frame among consecutive frames, decode, from a data stream, a quantized spectrum, a linear prediction coefficient based spectral envelope representation and a fundamental frequency related parameter. The decoder is configured to determine a spectral shaping function from the linear prediction coefficient based spectral envelope representation using a first manner below a pitch frequency determined from the fundamental frequency related parameter, and a second manner above the pitch frequency, to spectrally shape the quantized spectrum using the spectral shaping function to obtain a dequantized spectrum and to reconstruct the predetermined frame using the dequantized spectrum. Furthermore, the audio decoder is configured so that the spectral shaping function is, at a predetermined spectral position, lower if the pitch frequency is spectrally higher than the predetermined spectral position, than compared to if the pitch frequency is spectrally lower than the predetermined spectral position. Further decoders, encoders and methods are disclosed.

A method and apparatus for coding an audio signal are disclosed. According to one embodiment, the method for decoding an audio signal includes receiving a bitstream containing information about a first audio signal. The method further includes generating a second audio signal based on first quantization information and a first frequency spectrum obtained from the bitstream. The first quantization information may include quantization information generated based on a first complex polynomial having complex linear prediction coefficients (LPCs) corresponding to the first audio signal as coefficients.

Disclosed is an apparatus and method for audio encoding / decoding that is robust against coding distortion in a transition section. An audio encoding method includes outputting a frequency domain signal by time-to-frequency (T / F) transform of an input signal, outputting a frequency domain residual signal in which a frequency axis envelope is removed from the frequency domain signal by applying frequency domain noise shaping (FDNS) encoding to the frequency domain signal, outputting a time domain residual signal in which a time axis envelope is removed by performing linear prediction coefficient (LPC) analysis based on the frequency domain residual signal, and quantizing and transmitting the time domain residual signal.

Embodiments according to the invention comprise an audio decoder configured to, for a predetermined frame among consecutive frames, decode, from a data stream, a quantized spectrum and a linear prediction coefficient based envelope representation. Furthermore, the decoder is configured to locate, in the quantized spectrum, zero-quantized portions and non-zero-quantized portions and to derive a dequantized spectrum using in zero-quantized portions of the quantized spectrum, filling the quantized spectrum with a synthesized spectral data modified depending, according to a first manner, on the linear prediction coefficient based envelope representation, and in non-zero-quantized portions of the quantized spectrum, modifying the quantized spectrum depending, in a second manner, on the linear prediction coefficient based envelope representation. In addition, the decoder is configured to reconstruct the predetermined frame using the dequantized spectrum. The audio decoder is configured so that, for a predetermined portion, the modification according to the first manner and the modification according to the second manner cause a spectral quantization noise shaping which comprises different smoothness characteristics. Beyond that, corresponding encoders and methods are disclosed.

A coding method and a decoding method are provided which can use in combination a predictive coding and decoding method which is a coding and decoding method that can accurately express coefficients which are convertible into linear prediction coefficients with a small code amount and a coding and decoding method that can obtain correctly, by decoding, coefficients which are convertible into linear prediction coefficients of the present frame if a linear prediction coefficient code of the present frame is correctly input to a decoding device. A coding device includes: a predictive coding unit that obtains a first code by coding a differential vector formed of differentials between a vector of coefficients which are convertible into linear prediction coefficients of more than one order of the present frame and a prediction vector containing at least a predicted vector from a past frame, and obtains a quantization differential vector corresponding to the first code; and a non-predictive coding unit that generates a second code by coding a correction vector which is formed of differentials between the vector of the coefficients which are convertible into the linear prediction coefficients of more than one order of the present frame and the quantization differential vector or formed of some of elements of the differentials.

An embodiment according to the present invention includes an audio decoder configured to decode a quantized spectrum and an envelope representation based on linear prediction coefficients from a data stream for a predetermined frame of consecutive frames. Further, the decoder is configured to locate a zero-quantized portion and a non-zero quantized portion in the quantized spectrum and derive the quantized spectrum by filling the quantized spectrum in the zero-quantized portion of the quantized spectrum with synthetic spectrum data modified in accordance with an envelope representation based on linear prediction coefficients in a first manner, and modifying the quantized spectrum in the non-zero quantized spectrum portion in a second manner in accordance with an envelope representation based on linear prediction parameters. Further, a decoder is configured to reconstruct a predetermined frame using the dequantized spectrum. The audio decoder is configured such that, for the predetermined portion, the modification according to the first manner and the modification according to the second manner are such that the spectrally quantized noise shaping comprises different smoothness characteristics. In addition, a corresponding encoder and method are also disclosed.

A technology of accurately coding and decoding coefficients which are convertible into linear prediction coefficients even for a frame in which the spectrum variation is great while suppressing an increase in the code amount as a whole is provided. A coding device includes: a first coding unit that obtains a first code by coding coefficients which are convertible into linear prediction coefficients of more than one order; and a second coding unit that obtains a second code by coding at least quantization errors of the first coding unit if (A−1) an index Q commensurate with how high the peak-to-valley height of a spectral envelope is, the spectral envelope corresponding to the coefficients which are convertible into the linear prediction coefficients of more than one order, is larger than or equal to a predetermined threshold value Th1 and / or (B−1) an index Q′ commensurate with how short the peak-to-valley height of the spectral envelope is, is smaller than or equal to a predetermined threshold value Th1′.

The invention provides a snore detection method, device and equipment and a readable storage medium. The snore detection method comprises the steps that the Mel-frequency cepstral coefficient and the linear prediction coefficient of a microphone sound signal are calculated; performing feature splicing and dimension reduction processing on the Mel-frequency cepstrum coefficient and the linear prediction coefficient to obtain a target feature sequence; calculating a target feature vector corresponding to the micro-Doppler feature of the target part of the body of the user based on the radar echo signal; wherein the microphone sound signal and the radar echo signal are signals synchronously collected in a user sleep environment; the micro-Doppler features are used for reflecting the life activity state of the target part; and inputting the target feature sequence and the target feature vector into a preset fusion processing model to output a snore detection result. Through the implementation of the scheme of the invention, the feature data collected by the two sensors are fused, the snore recognition capability in a home scene can be effectively improved, and the complexity of data processing and the dependence of data processing on a network are reduced.

The invention discloses a multi-mode child snore monitoring system and detection system based on a mobile terminal, the mobile terminal comprises an audio acquisition module, a data storage module and a data processing module, the data storage module stores audio data acquired by the audio acquisition module, and the data processing module processes the audio data. The data processing module preprocesses the audio data, extracts a Mel-frequency cepstrum coefficient, a filter bank coefficient, a linear prediction coefficient and short-time average energy of each frame of audio data and stores the Mel-frequency cepstrum coefficient, the filter bank coefficient, the linear prediction coefficient and the short-time average energy in the data storage module, and an identification model is further configured in the data storage module; and the data processor module is controlled by a preset control signal, and calls the recognition model for calculation based on the Mel-frequency cepstrum coefficient, the filter bank coefficient, the linear prediction coefficient and the short-time average energy so as to output a child snore monitoring result.

An embodiment according to the present invention includes an audio decoder configured to decode, from a data stream, a quantized spectrum, a spectral envelope representation based on linear prediction coefficients, and a fundamental frequency related parameter for a predetermined frame of successive frames. The decoder is configured to determine a spectral shaping function from a spectral envelope representation based on linear prediction coefficients using a first manner below a fundamental tone frequency determined from the fundamental frequency related parameters and a second manner above the fundamental tone frequency, to spectrally shape the quantized spectrum using the spectral shaping function to obtain a dequantized spectrum, and to decode the dequantized spectrum using the dequantized spectrum. And reconstructing the predetermined frame using the dequantized spectrum. Furthermore, the audio decoder is configured such that, if the pitch frequency is spectrally higher than a predetermined spectral position, the spectral shaping function is lower at the predetermined spectral position than if the pitch frequency is spectrally lower than the predetermined spectral position. A decoder, an encoder, and a method are also disclosed.

This invention relates to an audio decoder for providing a decoded audio information on the basis of an encoded audio information comprising linear prediction coefficients (LPC), a respective method, a respective computer program for performing such a method and an audio signal for a storage medium having stored such an audio signal, the audio signal having been treated with such a method. The audio decoder comprises a tilt adjuster configured to adjust a tilt of a noise using linear prediction coefficients of a current frame to obtain a tilt information and a noise inserter configured to add the noise to the current frame in dependence on the tilt information obtained by the tilt calculator. Another audio decoder according to the invention comprises a noise level estimator configured to estimate a noise level for a current frame using a linear prediction coefficient of at least one previous frame to obtain a noise level information; and a noise inserter configured to add a noise to the current frame in dependence on the noise level information provided by the noise level estimator. Thus, side information about a background noise in the bitstream may be omitted.

The invention provides a control method of a liquid cooling system and electronic equipment, and can be applied to the technical field of heat dissipation. The control method of the liquid cooling system comprises the following steps: determining load grade information of electronic equipment in a target time period according to performance index data of the electronic equipment in the target time period; preprocessing the target data to obtain a plurality of sub-target data frames; performing linear predictive coding on the plurality of sub-target data frames to obtain a plurality of sub-linear predictive coefficients; based on the load level information, matching reference data from a preset health reference library; comparing the plurality of sub-target data frames and the plurality of sub-linear prediction coefficients with the matched reference data, and determining residual energy and coefficient distance corresponding to the target data; and determining a health index based on the residual energy and the coefficient distance corresponding to the target data, and generating early warning information based on the health index and a preset health threshold, the early warning information being used for prompting to perform a control operation corresponding to the early warning information on the liquid cooling system.

A coding method and a decoding method are provided which can use in combination a predictive coding and decoding method which is a coding and decoding method that can accurately express coefficients which are convertible into linear prediction coefficients with a small code amount and a coding and decoding method that can obtain correctly, by decoding, coefficients which are convertible into linear prediction coefficients of the present frame if a linear prediction coefficient code of the present frame is correctly input to a decoding device. A coding device includes: a predictive coding unit that obtains a first code by coding a differential vector formed of differentials between a vector of coefficients which are convertible into linear prediction coefficients of more than one order of the present frame and a prediction vector containing at least a predicted vector from a past frame, and obtains a quantization differential vector corresponding to the first code; and a non-predictive coding unit that generates a second code by coding a correction vector which is formed of differentials between the vector of the coefficients which are convertible into the linear prediction coefficients of more than one order of the present frame and the quantization differential vector or formed of some of elements of the differentials.

The invention provides a speaker classification method and device based on double-record data, and the method comprises the steps: obtaining historical double-record data in a business handling process, extracting audio data in the historical double-record data, extracting a Mel-frequency cepstral coefficient and a linear prediction coefficient based on the audio data, and carrying out the recognition of the Mel-frequency cepstral coefficient and the linear prediction coefficient; after feature fusion, training the speaker classification model to obtain an alternative classification model; and deploying the alternative model in a server, carrying out migration training based on recording data received in real time until the loss is smaller than a threshold value, and obtaining a target classification model for speaker classification and labeling of real-time double-recording data. The method solves the problem that in the prior art, the accuracy of a method for classifying speakers by analyzing double-recording data is insufficient.

This application provides a control method and electronic device for a liquid cooling system, which can be applied to the field of heat dissipation technology. The control method includes: determining the load level information of the electronic device within a target time period based on performance index data; preprocessing the target data to obtain multiple sub-target data frames; performing linear prediction encoding on each of the multiple sub-target data frames to obtain multiple sub-linear prediction coefficients; matching benchmark data from a preset health benchmark library based on the load level information; comparing the multiple sub-target data frames and multiple sub-linear prediction coefficients with the matched benchmark data to determine the residual energy and coefficient distance corresponding to the target data; determining health indicators based on the residual energy and coefficient distance corresponding to the target data; generating early warning information based on the health indicators and a preset health threshold; and using the early warning information to prompt control operations corresponding to the early warning information on the liquid cooling system.