238 results about "Arithmetic coding" patented technology

A system and method for cross-domain knowledge transfer in federated compression networks. The system enables efficient lossless data compression across diverse data types by intelligently sharing compression strategies between domains. A cross-domain knowledge transfer system identifies relationships between different data domains, adapts compression parameters accordingly, and optimizes learning processes to maximize knowledge reuse. The architecture may include a knowledge repository for storing domain features and compression patterns, domain mapping components that identify similarities, and transfer learning optimization that enables efficient adaptation with minimal examples. This approach significantly accelerates model training for new domains while improving compression performance. Applications include satellite telemetry systems where efficient compression is critical for transmitting large information sets between distant locations. The system may employ probability prediction driven arithmetic coding paired with long short-term memory networks, enhanced by cross-domain knowledge sharing that adapts successful compression strategies from one domain to another while preserving domain-specific optimization.

The invention relates to a multi-dimensional transmission data compression sensing method and system for urban emergency rescue, and the method comprises the steps: obtaining multi-modal data, carrying out the normalization processing and sparse representation, and obtaining the sparse coefficient of each modal; constructing a block diagonal random matrix for each modal sparse coefficient to perform independent compression sampling, generating a low-dimensional observation value before coding, and performing arithmetic coding and low-density parity check code error correction coding to generate an anti-interference code stream; and performing low-density parity check code decoding and arithmetic decoding on the anti-interference code stream to obtain a decoded low-dimensional observation value, performing iteration by using an orthogonal matching pursuit algorithm, recovering sparse coefficients of each mode, performing reconstruction to obtain text, image and audio features, and projecting the text, image and audio features to a shared semantic space to obtain a shared semantic space. And calculating a joint semantic weight through a cross-modal attention mechanism, and generating a multi-modal fusion result with consistent semantics. The method disclosed by the invention provides an efficient, robust and semantic collaborative end-to-end solution for urban emergency rescue.

The invention discloses a point cloud compression method and device based on a multi-scale octree attention mechanism, and relates to the field of image processing, and the method comprises the steps: an encoder network receives point cloud data, carries out the down-sampling and feature extraction of a point cloud through a downscaling feature extractor, obtains a downscaled deep feature point cloud, and carries out the feature extraction of the point cloud; the method comprises the following steps of: firstly, encoding the data into an octree in a recursive manner, constructing a context window according to a relationship among octree nodes, introducing a multi-head attention mechanism to perform feature fusion on the octree nodes to obtain an occupancy probability of the octree nodes, and compressing the occupancy probability into a bit stream by using arithmetic encoding; and the decoder network decompresses the bit stream to obtain a reconstructed point cloud, upsampling and feature reconstruction are performed on the reconstructed point cloud by using an upscaling feature reconstruction device, and finally a reconstructed point cloud with the same resolution as the initial point cloud is obtained. According to the invention, on the premise of ensuring the quality of the same point cloud, the point cloud compression efficiency is effectively improved, and the bit overhead is reduced.

One example method includes receiving, by a large language model (LLM), input text to be compressed, defining a size of a rolling window of previous tokens, generated prior to receipt of the input text, that the LLM is permitted to consider in a conditional probability estimate, generating, by the LLM, tokenized text based on the input text, and the tokenized text comprises a sequence of tokens, based on the previous tokens, obtaining a probability mass function of a next token of the sequence, providing the probability mass function as an input to an arithmetic coding (AC) scheme, and assigning, by the AC scheme, a respective binary code to the token with a highest probability as assigned by the LLM.

A method performed by at least one processor of a video decoder includes receiving a coded video bitstream including at least one picture and one or more syntax elements encoded in accordance with multi-hypothesis arithmetic coding. The method further includes decoding each syntax element from the one or more syntax elements based on the multi-hypothesis arithmetic coding. The method further includes selecting a probability update rate from a plurality of probability update rates based on a predetermined condition, the plurality of probability update rates including a first probability update rate that is higher than a second probability update rate. The method further includes updating at least one probability model utilized in the multi-hypothesis arithmetic coding based on the selected probability update rate. The method further includes decoding at least one block in the at least one picture based on the decoded one or more syntax elements.

This disclosure provides methods, devices and systems for encoding data, to achieve a symbol distribution, for wireless communication. One implementation includes a method in which arithmetic coding (AC) encoding is constrained by a set of one or more target compositions for symbol sequences. The target compositions are known, and the encoding method is performed in multiple iterations. Each iteration generates a symbol and establishes a composition prefix, which is taken into account in the subsequent iteration. The methods generate output sequences defining symbols that are used to encode data for transmission.

In one implementation, a method of encoding or decoding point cloud data is provided, comprising: obtaining a feature map representing attributes of voxels in an octree structure; determining one or more probability distribution parameters for a probability density function associated with an attribute of a current voxel, based on the feature map; determining a probability mass function of the attribute of the current voxel based on the one or more probability distribution parameters for the probability density function for the current voxel; and encoding or decoding attribute information of the current voxel in the octree structure, based on the probability mass function of the attribute for the current voxel.

A system and method for federated two-stage compression with federated joint learning. The system and method proposed allow for fast and efficient lossless data compression of a large variety of data types. The system and method have a variety of real-world applications, including deep learning solutions for telemetry, tracking, and command subsystems for satellites. Satellites and their control centers are incredibly spaced apart which makes data compression an extremely important process to transmit large sets of information in a low-latency, high-efficiency environment. The proposed system and method utilize probability prediction driven arithmetic coding which provides faster encoding times and higher compression ratios when paired with a long short-term memory system for data compression.

The invention is suitable for the technical field of image processing, and provides an FPGA-based cache-free real-time image compression transmission method and system, and the method comprises the steps: sequentially carrying out the block processing, level translation processing and component transformation processing of image data, and obtaining a preprocessed image; carrying out discrete wavelet transform processing on the preprocessed image to obtain a wavelet output result, and storing the wavelet output result in an FIFO (First In First Out) memory; for the wavelet output result, obtaining context information and decision bit information by adopting a bit-plane coding method of four-channel parallel and all-bit-plane parallel processing; carrying out arithmetic coding on the context information and the decision bit information by adopting a multi-stage pipeline processing method to obtain a compressed code stream; and converting the compressed code stream into a standardized code stream format. According to the method, the memory resource demand can be effectively reduced, the delay is reduced, and the real-time image compression requirement is met.

The invention relates to a task-oriented semantic communication-oriented adaptive source channel coding system, which belongs to the field of wireless communication, and comprises a sending end for extracting a semantic feature map for evaluation from input data through a feature extractor initialized based on a pre-training large model, executing nested quantization and arithmetic coding based on a calculated importance score, and sending the semantic feature map to a task-oriented semantic communication-oriented adaptive source channel coding system; generating a scalable semantic bit stream with importance levels; the physical layer is used for carrying out reordering and grouping according to importance levels of bit streams, a channel encoder is combined with an interleaver to realize unequal error protection of semantic matching, and encoded bits are mapped to sign bits with different fault-tolerant capabilities in a constellation diagram through division modulation; and a receiving end: recovering the semantic feature map through arithmetic decoding and nested inverse quantization. According to the invention, on the premise of keeping compatibility with a digital communication protocol stack, through distortion-driven joint rate self-adaption, the transmission effectiveness for machine tasks and human vision is obviously improved under a limited bandwidth.

A method and an apparatus including processing circuitry are provided. The processing circuitry determines previous probability information associated with each entropy coded region of multiple regions in a previous picture. The processing circuitry selects, based on (i) a location or a quantization parameter of an independently decodable coding segment in a current picture or (ii) syntax information of the independently decodable coding segment, a region among the multiple regions. Initial probability information for one or more current syntax elements of a block in the independently decodable coding segment is determined based on the previous probability information associated with the selected region. The one or more current syntax elements are first to be entropy decoded in the independently decodable coding segment. The processing circuitry entropy decodes coded bits associated with the one or more current syntax elements into a bin string based on the initial probability information.

Devices, systems, and methods for digital video coding, which includes matrix-based intra prediction methods for video coding, are described. In a representative aspect, a method for video processing includes encoding a current video block of a video using a matrix intra prediction (MIP) mode in which a prediction block of the current video block is determined by performing, on previously coded samples of the video, a boundary downsampling operation, followed by a matrix vector multiplication operation, and selectively followed by an upsampling operation; and adding, to a coded representation of the current video block, a syntax element indicative of applicability of the MIP mode to the current video block using arithmetic coding in which a context for the syntax element is derived based on a rule.

A device may obtain a first value for a first binarization symbol associated with video data and a second value for a second binarization symbol associated with the video data. The device may obtain a probability value associated with a third binarization symbol associated with the video data based on the first value, the second value, and a probability value associated with the first binarization symbol. The device may then decode the video data based on the probability value associated with the third binarization symbol. The first, second and third binarization symbols may be for a transform coefficient. In examples, the first, second and third binarization symbols may be associated with a context for entropy decoding. For example, the first, second and third binarization symbols may be associated with an entropy decoding context for decoding the video data using context-based adaptive binary arithmetic coding (CABAC).

The invention relates to the technical field of industrial data compression, in particular to an industrial operation and maintenance data hybrid compression storage method, system and device and a medium. High compression rate and lossless storage are realized by distinguishing time sequence data from non-time sequence feature vectors and respectively adopting Huffman coding and LLM-driven arithmetic coding; and through efficient compression processing, especially effective reduction of feature vector redundancy, the space required by data storage is greatly reduced, so that the storage cost of the operation and maintenance data of the generic semiconductor industrial equipment is reduced.

Methods, apparatus, and computer readable storage medium for multi-symbol arithmetic coding in video decoding. The method includes receiving a coded video bitstream; obtaining an array of cumulative distribution functions and corresponding M symbols, wherein M is an integer larger than 1; and performing, based on the array of cumulative distribution functions, arithmetic decoding on the coded video bitstream to extract at least one symbol by: for each extracted symbol, updating the array of cumulative distribution functions according to the extracted symbol based on at least one probability update rate, and continuing, based on the updated array of cumulative distribution functions, arithmetic decoding on the coded video bitstream to extract next symbol.

The invention relates to the technical field of point cloud, in particular to a coding and decoding method for point cloud compression. The invention provides a coding method for geometric compression of a point cloud, which comprises the following steps of: performing layer-by-layer down-sampling and layer-by-layer up-sampling on each spatial scale of a point cloud group, and training a network through over-fitting to obtain optimized network parameters; based on the network parameters, performing layer-by-layer up-sampling on each spatial scale of the point cloud group, predicting the occupancy probability of the child nodes and taking the occupancy probability as a code table, performing arithmetic coding on occupancy codes of the child nodes, and generating an occupancy code stream; and coding the network parameters to generate a network parameter code stream. According to the invention, the adjacent frames with similar features in the point cloud sequence share the lightweight network parameters for coding and decoding, so that the repeated transmission of the network parameters is reduced, and the average bandwidth consumption of the network parameters in the transmission process is reduced.

The invention discloses a high-definition video low-delay compression method based on self-adaptive adjustment, and aims to solve the problems of insufficient self-adaptive adjustment capability, too high encoding and decoding delay and low code rate and network bandwidth matching degree in the existing high-definition video compression technology. The method comprises the following steps: firstly, acquiring a high-definition video stream and a real-time network bandwidth parameter, extracting intra-frame texture complexity, inter-frame motion intensity and scene switching frequency characteristics, and constructing a two-dimensional adaptive compression strategy in combination with a real-time bandwidth; low-delay coding is executed by adopting parallel frame fragmentation and simplified adaptive arithmetic coding, adaptive code rate control is realized by combining code rate overflow early warning and intra-frame key region priority bit allocation, and finally, fast inverse transformation, quantization recovery and motion compensation optimization are completed at a decoding end based on strategy parameters synchronously transmitted at a coding end so as to reconstruct a high-definition video frame. According to the invention, the adaptive adjustment capability of video compression is improved, low time delay of coding and decoding is realized, and the method is suitable for real-time high-definition video transmission scenes such as remote monitoring and video conferences.

An encoder includes circuitry and memory coupled to the circuitry. In residual coding of a current block to which an orthogonal transform is not applied, when coefficient information flags relating to a coefficient in the current block are encoded, the circuitry: performs a level mapping process to transform the coefficient to a second coefficient by using a predicted value, in which the predicted value is determined based on neighboring coefficients of the coefficient within the current block; encodes second coefficient information flags by Context-based Adaptive Binary Arithmetic Coding (CABAC), each of the second coefficient information flags relating to the second coefficient; and encodes a remainder value of the coefficient with Golomb-Rice code, and when the coefficient information flags are not encoded, in operation, the circuitry: skips the level mapping process; and encodes a value of the coefficient with the Golomb-Rice code.

To provide a data compression system that combines large-scale language models (LLMs) with distributed computing to solve the problems that conventional compression techniques only consider statistical redundancy and are unable to exploit semantic structure or contextual information, and furthermore, large-scale model-based compression requires high resources and fails to fully exploit the interrelationships in multimodal data.SOLUTION: The present invention achieves highly efficient compression even in resource-constrained environments, by utilizing the interrelationships between different data formats using a multimodal understanding module, selecting an optimal model according to the data characteristics and communication environment using an adaptive model selection module, utilizing shared knowledge between terminals via a distributed collaborative compression module, transferring the compression capabilities of large-scale models using a knowledge distillation module to lightweight models, and integrating efficient coding using an adaptive arithmetic coding module.SELECTED DRAWING: None

An apparatus with a processor receives a bitstream including a syntax element indicating whether prediction errors, mesh connectivity information, and additional basemesh syntax elements are transmitted in a single entropy packet or a plurality of entropy packets, a first set of arithmetically coded prediction errors, and a second set of arithmetically coded prediction errors for decoding a mesh frame. The processor decodes the syntax element and arithmetically decodes the first and second set of arithmetically coded prediction errors based on the syntax element. In one example, the first and second set of arithmetically coded prediction errors are included in a single entropy packet if the syntax element has a first value and the first set of arithmetically coded prediction errors is in a first entropy packet and the second set of arithmetically coded error predictions is in a second entropy packet if the syntax element has a second value.

An encoding device (100) is provided with: a memory (152); and a circuit (151) capable of accessing the memory (152), the circuit (151), during operation, encodes a connection type pertaining to a connection relationship between a surface to be processed and an unprocessed surface by context-based arithmetic encoding for each of a plurality of surfaces constituting a three-dimensional grid, and, during encoding of a current connection type, which is the connection type to be encoded, encodes the current connection type to be encoded by context-based arithmetic encoding for each of the plurality of surfaces constituting the three-dimensional grid. The context to be applied to the arithmetic encoding is selected from a plurality of contexts using a previous connection type that is the connection type encoded prior to the current connection type.

This invention discloses a dual-stream point cloud compression method and system based on semantic scene graphs, relating to the field of point cloud compression. It constructs a point cloud compression network comprising an encoder network and a decoder network: In the encoder network, semantic labels are mapped onto the input point cloud, and scene graphs and octree representations are constructed respectively. Semantic modulation parameters are generated based on the scene graph and encoded to obtain a semantic bitstream. The octree representation, under the influence of semantic modulation parameters, undergoes feature enhancement and attention probability prediction to obtain node occupancy probabilities, and a geometric bitstream is generated through arithmetic encoding. These two bitstreams serve as compressed data. In the decoder network, the bitstreams are decoded, and the octree structure and scene graph information are reconstructed. Node feature representations are recovered by combining the semantic modulation parameters. The reconstructed octree structure is mapped back to the point cloud space through spatial querying and indexing to reconstruct the point cloud. This invention solves the problem of local geometric distortion in low-bitrate octree point cloud compression, ensuring reconstruction quality while reducing the transmission bitrate.

Methods and apparatuses are provided to encode data into a bitstream and to decode data from a bitstream. According to the method, the length of the bitstream can be reduced by including only the relevant significant bits within the trailing bits of the encoding process. The amount of these leading trailing bits is determined and the trailing bits, which have the least amount, can be constructed. Indications of the amount of leading trailing bits are included into the bitstream. Therefore, padding is not required, resulting in less bits, which need to be signaled.

The invention proposes a method and a device for arithmetic encoding of a current spectral coefficient using preceding spectral coefficients. Said preceding spectral coefficients are already encoded and both, said preceding and current spectral coefficients, are comprised in one or more quantized spectra resulting from quantizing time-frequency-transform of video, audio or speech signal sample values. Said method comprises processing the preceding spectral coefficients, using the processed preceding spectral coefficients for determining a context class being one of at least two different context classes, using the determined context class and a mapping from the at least two different context classes to at least two different probability density functions for determining the probability density function, and arithmetic encoding the current spectral coefficient based on the determined probability density function wherein processing the preceding spectral coefficients comprises non-uniformly quantizing absolutes of the preceding spectral coefficients for use in determining of the context class.

Methods and apparatus are provided for initializing a set of context model probability sets for slices in context-based adaptive binary arithmetic coding (CABAC). An exemplary video decoding method includes selecting a first set of probability parameters for initializing one or more context models of a B-slice from a predefined plurality of sets of probability parameters; and performing entropy decoding of the B-slice based on the one or more context models and the first set of probability parameters, where the selection is based on a coding condition of the B-slice or a signal in the bitstream.

The present disclosure relates to a method and apparatus for compressing parameters of a neural network, and specifically describes an encoder for encoding weight parameters of a neural network. The encoder is configured to: obtain a plurality of weight parameters of the neural network; encoding a weight parameter of the neural network using context-dependent arithmetic coding; selecting a context for encoding a weight parameter or for encoding a syntax element of a digital representation of a weight parameter in accordance with one or more previously encoded weight parameters and / or one or more previously encoded syntax elements in accordance with the digital representation of the one or more weight parameters; and encoding the weight parameter or the syntax element of the weight parameter using the selected context. A corresponding decoder, quantizer, method and computer program are also described.

The application relates to the technical field of data compression storage, and provides a high-throughput genome sequencing quality score data parallel compression method.The method comprises the following steps: segmenting an original gene sequencing file; performing random sampling and k-mer analysis on the sampling data to obtain statistical characteristic information and establish a parallel sequence partition model for binary classification; splicing two partition files obtained through binary classification according to splicing parameters; predicting the to-be-compressed file through a multivariate linear regression analysis prediction method to obtain compression rate gain and establish a parallel four-level run prediction mapping model for data redundancy elimination; and performing context modeling on two redundancy-eliminated subfiles through a multi-core processor cluster, and performing cascade compression in combination with arithmetic coding to obtain a final compressed file.Under the premise of significantly reducing quality score data compression time and peak memory overhead, the application also improves the quality score data compression rate, reduces the size of to-be-compressed storage files, and saves the construction cost of basic storage facilities.

The invention proposes a method and a device for arithmetic encoding of a current spectral coefficient using preceding spectral coefficients. Said preceding spectral coefficients are already encoded and both, said preceding and current spectral coefficients, are comprised in one or more quantized spectra resulting from quantizing time-frequency-transform of video, audio or speech signal sample values.Said method comprises processing the preceding spectral coefficients, using the processed preceding spectral coefficients for determining a context class being one of at least two different context classes, using the determined context class and a mapping from the at least two different context classes to at least two different probability density functions for determining the probability density function, and arithmetic encoding the current spectral coefficient based on the determined probability density function wherein processing the preceding spectral coefficients comprises non-uniformly quantizing absolutes of the preceding spectral coefficients for use in determining of the context class.

An encoding device (100) is provided with a memory (152) and a circuit (151) capable of accessing the memory (152), the circuit (151), during operation, encodes a connection type by context-based arithmetic encoding (S1410) for each of a plurality of surfaces constituting a three-dimensional grid, selects a context (S1411) using a first previous connection type during encoding of a current connection type, and performs encoding of the current connection type using a second previous connection type during encoding of the current connection type. When the first previous connection type is shifted by the angle type, the context is selected according to the number of more than one angle type continuously encoded up to the first previous connection type.

Aspects of the present disclosure generally relate to wireless communications. In some aspects, a transmitting node may obtain a k-bit sequence of information bits. The transmitting node may encode the k-bit sequence into an output sequence corresponding to a length n symbol sequence of a set of symbol sequences of length n on a symbol set Am according to a first stage of energy-based arithmetic coding for probability amplitude shaping (PAS) and a second stage of energy-based arithmetic coding for PAS. The transmitting node may perform a transmission to a receiving node based at least in part on the n-length symbol sequence. Numerous other aspects are described.