283 results about "Entropy encoding" patented technology

In one implementation, we propose a scalable framework where a base layer uses NN-based methods to compress the content for computer vision machine tasks and enhancement layer(s) use traditional predictive coding for human viewing. Typically, the based layer performs NN-based analysis to generate a latent tensor, which is entropy coded to produce the base layer bitstream. By performing synthesis on the latent tensor, an inter-layer predictor can be obtained for the enhancement layer(s). Since many machine tasks are not required to be performed for each frame, the base layer may skip analysis for some frames. The synthesis may be performed at the base layer or the enhancement layer(s). In one example, the base layer compresses features optimized for a machine task and the enhancement layer(s) rely on predictive coding. In another example, the enhancement layer(s) can use traditional scalable video compression methods.

The invention provides a lossless compression method and decompression method for power grid waveform data, equipment, a medium and a program product, and the method comprises the steps: carrying out the periodic feature analysis of the original sampling data of a power grid waveform, so as to determine the periodic structure of the original sampling data, and carrying out the segmentation processing of the original sampling data based on the periodic structure; for the sampling data in each segment, extracting differential information between adjacent sampling points to form a differential data sequence; dynamically determining a corresponding quantization parameter based on the statistical distribution characteristic of the differential data sequence, and performing adaptive quantization processing on the differential data sequence to generate quantized differential data; and entropy coding processing is carried out on the quantized differential data to generate compressed data used for representing the original sampling data, and characteristic parameters used for reconstructing the original sampling data are stored in the compressed data in an associated mode. Unification of high compression ratio, low power consumption and strict lossless reconstruction is realized, and the application requirements of long-term monitoring and remote transmission of the edge side can be met.

The invention discloses a millimeter wave radar point cloud compression method based on heterogeneous representation and an implicit neural network. Firstly, point cloud fine geometric features are extracted through a point-based compression network, global semantics are aggregated through a voxel-based compression network, and point-voxel heterogeneous joint representation and preliminary compression are realized. Secondly, constructing an adaptive octree to code a compressed data space structure to generate a partitioned bit stream, and performing statistical compression on the voxel-level high-dimensional features by using context-aware entropy coding to generate a feature bit stream; then, joint transmission is performed on the partition bit stream, the feature bit stream and the neural network weight for implicit representation. And finally, decoding and recovering the coarse-grained point cloud at a receiving end, inputting the coarse-grained point cloud and a neural network weight into a point-based synthesis network based on implicit neural representation, and finally reconstructing a high-fidelity point cloud through continuous surface modeling and detail compensation. According to the method, the compression efficiency and the reconstruction quality of the millimeter wave radar point cloud under the low code rate are remarkably improved.

The invention discloses a multi-modal feature adaptive fusion and compression semantic communication and classification method and device, and the method comprises the steps: carrying out the multi-modal feature extraction of a collected multi-modal signal, and obtaining a multi-modal feature; carrying out adaptive fusion on the multi-modal features to obtain fused features; and entropy coding compression is performed on the fused features to obtain fused compressed features, and the fused compressed features are converted into binary code streams through an encoder-decoder mechanism for realizing semantic communication with a task module. The method has the advantages of being high in fusion efficiency, friendly in compression, low in calculation overhead, small in communication cost and bandwidth occupation and the like, and the dual requirements for intelligent sensing and task execution capacity under the condition that the bandwidth is limited or the computing power is limited are met.

Systems, apparatuses, methods, and computer-readable media are described for coding RAHT coefficients based on coefficients of neighboring nodes. An encoder may use a forward RAHT transform to obtain all RAHT coefficients and may entropy encode all coefficients in the bitstream. A decoder may decode all coefficients and then backward RAHT transform the decoded coefficients to obtain attributes. A coder (e.g., encoder or decoder) may use already-coded coefficients of neighboring RAHT nodes to select a context and entropy code coefficients of a current RAHT node based on the selected context.

The invention relates to the technical field of virtual reality, and discloses a touch display screen panoramic interaction method based on virtual reality. The method comprises the following steps: deploying a tactile sensing network on the surface of a touch display screen, collecting space coordinates and pressure data of a touch point in real time, and synchronously obtaining a computing resource utilization rate and network bandwidth data of a virtual reality processing node; classifying the touch data to generate behavior mode labels, and determining an optimal touch point distribution set by using a dynamic load balancing algorithm in combination with resource data; performing normalization and entropy coding compression on the original interaction data stream to generate a standardized interaction data packet; transmitting to a temporary buffer area of the distributed database, activating a persistent trigger, acquiring a storage node state through a log synchronization module, and determining an optimal path for fragmentation storage. According to the method, multi-dimensional data are integrated, resource allocation and data processing storage are optimized, and the real-time performance and stability of virtual reality interaction are enhanced.

The invention provides an immersive video coding method and system based on 3DGS, and the method comprises the steps: carrying out the sparse reconstruction of each frame of multi-view video data, and obtaining an initial point cloud; determining anchor points of three-dimensional Gaussian distribution in each frame; extracting spatial context information of the anchor points through multi-resolution hash coding, and splicing the spatial context information with the features of the anchor points to form fusion features; predicting parameters of three-dimensional Gaussian distribution corresponding to each anchor point through a neural network by using the fusion features and camera parameters, and obtaining a rendered image; calculating the color loss between the rendered image and the original video frame, and optimizing the parameters of the anchor points; performing quantization and entropy coding on the optimized parameters of the anchor points; in combination with the color loss and the coding rate, performing rate distortion optimization on the quantization parameter to generate compressed three-dimensional scene representation; and repeating the steps for each frame of the multi-view video, and finally outputting a compressed immersive video code stream. According to the invention, high-quality representation and efficient compression of the immersive video are realized, and the rate-distortion performance is improved.

In end-to-end compression, a deep neural-network based encoder can be used to encode an image. The embeddings output from the encoder are quantized and encoded with a lossless encoder. Advantageously, at least one embodiment allows improving the latent entropy coding by further reducing the redundancies in the quantized latent. To that end, at least one embodiment discloses taking into account channels importance by coding an indication of a channel activity (or significance): performing post-conditional entropy coding by computing conditional probability based on a context afterwards: using channels reordering to improve inter channel correlation: or performing RDOQ like process by optimizing the main latent for a particular image.

The invention discloses a deep learning image compression method and system based on wavelet domain double branches, and the method comprises the steps: carrying out the coding and decoding through a deep neural network in combination with the wavelet domain features of high-frequency and low-frequency double branches, carrying out the coding and mapping of an original image to a compact potential feature, carrying out the super-prior coding, quantization and entropy coding, so as to generate a code stream, and carrying out the compression of a deep learning image. Global context information is obtained through hyper-prior decoding, channel division is performed on the compact potential features, and context modeling based on space and channels is performed on each channel block by using the global context information so as to predict a mean value and a standard deviation of the channel blocks obeying Gaussian distribution, and the mean value and the standard deviation are used for guiding quantization and entropy coding of the channel blocks; and generating a code stream after image compression, mapping the potential features obtained by decoding back to the reconstructed image, and constructing rate distortion loss based on the control code rate of the potential features after decoding, the control code rate of super-prior decoding and the distortion of the original image and the reconstructed image so as to train a deep neural network for image compression.

The invention provides an intelligent compression transmission method and system for a message body, and the method achieves the efficient compression and intelligent transmission of a structured message through the technical means of structured data templated compression, type-aware entropy coding, network quality real-time monitoring, dynamic compression strategy decision and the like. The compression strategy can be adaptively adjusted according to the dynamic change of the network environment, and the data transmission efficiency is improved.

A method and an apparatus for encoding or decoding a video. An initial value for at least one parameter of a context of an entropy coder is determined. The context is associated to at least one binary symbol of a sequence of binary symbols to entropy code. The initial value is determined based on a bitrate determination for entropy encoding a set of binary symbols or based on decoding of an information representative of the initial value. The at least one parameter is initialized using the determined the initial value, and the sequence of binary symbols is entropy encoded or decoded based on the at least one parameter initialized.

Embodiments of the present disclosure relate to methods, devices, apparatuses, and program products for image encoding and decoding. An example method comprises: dividing an image into a plurality of channels; encoding a first channel of the plurality of channels based on entropy encoding; obtaining a first entropy parameter for the first channel; determining a second entropy parameter for a second channel of the plurality of the channels based on the first entropy parameter; and encoding the second channel based on the second entropy parameter. The embodiments of the present disclosure can significantly accelerate decoding of images.

Encoding or decoding syntax information associated with video information can involve identifying a coding context associated with a syntax element of a current coding unit of the video information, wherein the identifying occurs without using a syntax element of a neighboring block, and encoding or decoding the syntax element of the current coding unit based on the coding context.

The present application relates to the technical field of data compression, in particular to a data compression method and system, comprising: an LZ module, a text stream module, a sequence stream module, an interval entropy encoding framework, the LZ module is used for screening repeated data of original data, generating a text stream and a sequence stream; the LZ module screens repeated data through a hash verification method, specifically: calculating a hash value for the input data block, determining repeated data based on hash value comparison to avoid direct data content comparison; the text stream module is used for entropy encoding the text stream; the present application innovatively proposes an interval entropy encoding framework, constructs a more optimal probability model through artificial intelligence and mathematical methods, trains a more optimal finite state machine, and uses the optimal state machine in the compression and decompression process, efficiently solving the problem of slow decompression speed of neural network type methods.

The invention discloses a video processing method and device, equipment, a medium and a product, and the method comprises the steps: obtaining a to-be-coded video, carrying out the semantic feature extraction of the to-be-coded video through an encoder, and obtaining a semantic feature sequence; on the basis of at least two time windows with different sizes from large to small, carrying out sliding type common semantic extraction and quantification processing on the semantic feature sequence step by step to obtain a time quantification result and a time residual error; for each video frame in the same time window, carrying out sliding type common semantic extraction and quantization processing step by step based on at least two space windows with different sizes from large to small to obtain a space quantization result; and entropy coding is carried out on the time quantization result and the space quantization result to obtain a compressed code stream, so that the compression rate of the video can be improved, the memory occupied by the compressed video can be reduced, and the quality of the reconstructed video can be improved.

A method of encoding point cloud data includes storing, in a buffer, a maximum coordinate of a pair of coordinates of a applicable node, wherein the applicable node is a most recently encoded node having a same position as a current node along an applicable axis, and the pair of coordinates is for an axis different from the applicable axis; determining a context of a planar mode plane position of the current node, wherein determining the context of the planar mode plane position includes determining a distance value based on the maximum coordinate of the pair of coordinates of the applicable node, the distance value representing a distance between the current node and the applicable node; determining an increment value, the increment value indicating whether the distance value is greater than a threshold value; and determining a context index based on the increment value; and entropy encoding the planar mode plane position using the determined context.

The embodiment of the application discloses a kind of based on wavefront parallel processing to the method, device, storage medium and computer equipment of image H.264 encoding, it is related to image processing field.The application is processed adjacent macroblock line by double encoder parallelly, and respectively first buffer and second buffer are written to encoding bit stream.In macroblock line encoding completion, first buffer remaining bit stream and entropy encoding remaining bit stream are merged in real time, if the length after merging is less than 8 integer multiples, then appropriate bit stream is dynamically taken from the head of second buffer and spliced, complete code stream in line with H.264 standard is generated and outputed.After encoding is completed, the minimum splicing length is used as the offset of second buffer, continue to output byte stream, and the encoding of next macroblock line is started.The application only needs double macroblock line buffer, significantly improves the encoding speed, reduces memory occupation, and does not need to modify decoder, and compatibility is strong.

Methods and apparatuses are described for entropy encoding and decoding of a latent tensor, which includes separating the latent tensor into segments in the spatial dimensions, each segment including at least one latent tensor element. An arrangement of the segments is processed by a neural network; the neural network includes at least one attention layer. Based on the processed segment a probability model is obtained for entropy encoding or decoding of a latent tensor element.

The invention discloses a hardware acceleration device and method for high-throughput lossless data compression, and relates to the field of integrated circuit design. The device comprises a front-end LZ77 compression module, a middle buffer module and a rear-end entropy coding module, wherein the front-end LZ77 compression module is configured to search and match input data; the intermediate buffer module is configured to realize decoupling of front and rear end processing time domains through an asynchronous symbol buffer; the back-end entropy coding module is configured to receive an output of the intermediate buffer module, construct a microcode packet intermediate representation containing codeword information, and parse the microcode packet to generate a serial compressed bit stream. According to the method, invalid memory reading and assembly line pause can be effectively eliminated, the throughput and the energy efficiency ratio of the hardware compressor are remarkably improved on the premise of not sacrificing the compression rate, and the method is suitable for unloading acceleration of a high-performance solid state disk controller, an intelligent network card and a cloud server.

The present application discloses a VVC multi-directional motion refinement optimization method based on a Merge mode from the VVC inter prediction with a Merge process with a motion vector difference, which only has a small part of coding units (CUs) to search in a large step, and fails to consider the multi-directionality of the motion in the actual process. First, the first two MV information in the extended Merge candidate list is selected as the initial MV, and is expanded from four step lengths and eight directions; it is judged whether the current step length index is greater than 1, if yes, skip the chroma block motion compensation and the bi-directional optical flow prediction mode; otherwise, skip the chroma block motion compensation, and after the optimal Merge mode is obtained, entropy coding is carried out. Compared with the VVC coding standard, the average bit rate of Y, Cb and Cr is saved by 0.57%, 0.62% and 0.25% respectively by the VVC multi-directional motion refinement optimization method based on the Merge mode.

The application provides a lightweight variable bit rate multi-view image compression method and system, and relates to the technical field of image compression. The method comprises: downsampling, feature extraction and feature scaling of a single-view image to obtain a latent representation corresponding to a target bit rate; quantization and lossless entropy coding of the latent representation to obtain a final compressed bit stream; subsequent lossless entropy decoding and inverse scaling to restore the latent representation; feature fusion and upsampling of the restored latent representations of different views to generate a reconstructed compressed image. The model comprises a main encoder, a feature scaling module, a quantization module, an autoregressive entropy model, an arithmetic encoder, an arithmetic decoder, a feature inverse scaling module and a decoder. The application can efficiently compress image data, reduce computational complexity and storage space occupation while preserving image details and quality, thereby providing faster speed and lower bandwidth requirement for image transmission and storage.

Implementations of the disclosure provide a video decoding apparatus and method for transform coefficient sign prediction on a video decoder side. The method may include receiving a bitstream including a sequence of sign signaling bits for a set of candidate transform coefficients. The method may further include generating a set of predicted signs for the set of candidate transform coefficients associated with a transform block of a video frame from a video. The method may also include decoding the sequence of sign signaling bits based on one or more contexts used to entropy-encode the sequence of sign signaling bits to obtain an indication of correctness of the predicted signs of the respective candidate transform coefficients. The method may additionally include estimating original signs for the set of candidate transform coefficients based on the set of predicted signs and the decoded sequence of sign signaling bits.

A system for operating system virtualization using codebook encoding to improve entropy encoding methods. Training data sets are analyzed to determine occurrence frequency of each sourceblock in the data sets. A mismatch probability estimate is calculated at which any given data sourceblock received will not have a codeword in the codebook. Entropy encoding is used to generate codebooks comprising codewords for sourceblocks based on the frequency of occurrence of each. A “mismatch codeword” is inserted into the codebook to represent those cases when a block of data does not have a codeword.

The invention relates to the technical field of data compression, in particular to a real-time lossless compression method for linear scanning panoramic infrared image data. The method comprises the following steps: receiving current column original pixel data output by a linear column scanning infrared sensor; taking the original pixel data of the current column and the coded pixel data of the adjacent column in one frame as reference columns, and performing backward inter-column difference prediction to obtain predicted residual data of the current column; mapping the predicted residual data into a symbol value; carrying out entropy coding on the symbol value by using a static Huffman coding table which is pre-generated and solidified based on the statistical characteristics of the linear scanning infrared image, and outputting a compressed code stream; according to the scheme, the technical problem that high real-time performance and high compression ratio are difficult to consider when real-time lossless compression is carried out on linear scanning panoramic infrared image data in the prior art is solved.

The application discloses a millimeter wave radar point cloud compression method based on heterogeneous representation and implicit neural network. First, the point cloud fine geometric features are extracted through a point-based compression network, and then the global semantics are aggregated through a voxel-based compression network to realize point-voxel heterogeneous joint representation and preliminary compression. Secondly, an adaptive octree is constructed to encode the compressed data space structure to generate a partition bit stream, and a context-aware entropy coding is used to statistically compress the voxel-level high-dimensional features to generate a feature bit stream. Then, the partition bit stream, the feature bit stream and the neural network weight for implicit representation are jointly transmitted. Finally, the coarse-grained point cloud is decoded and recovered at the receiving end, and is input into a point-based synthesis network based on implicit neural representation together with the neural network weight, and through continuous surface modeling and detail compensation, a high-fidelity point cloud is finally reconstructed. The method significantly improves the compression efficiency and reconstruction quality of the millimeter wave radar point cloud at a low code rate.

The invention relates to a non-standard JPEG (Joint Photographic Experts Group) coding processing method for dynamic code rate control. The method comprises the following steps: presetting a multi-frame joint image processing scene, obtaining an original reference image, and carrying out combined processing of boundary expansion and random bias addition to obtain a preprocessed reference image; dividing macro blocks for the preprocessed reference image, and distributing an independent coding parameter configuration space after verification and compliance; frequency domain conversion is executed according to a division result, an initial quantization parameter is calculated through an index model, and a dynamic quantization table is dynamically adjusted; executing non-standard entropy coding to obtain an indexed macro block coding code stream; combining the indexed code stream and a macro block division result to construct an index table containing a macro block serial number, an initial address and a code stream length; analyzing the code stream to obtain quantized data, and performing inverse quantization by using a dynamic quantization table to obtain spatial domain data; enabling the format of the reconstructed image to be consistent with that of the preprocessed reference image through reverse preprocessing, and performing real-time processing on the linkage input image to obtain a reconstructed reference image; and non-standard JPEG coding processing of dynamic code rate control is realized.

A method and apparatus for entropy encoding / decoding a coefficient level are provided. A method for decoding a coefficient included in image data is provided, wherein the method comprises the steps of: determining a Rice parameter for a current coefficient based on a reference level of the current coefficient; parsing, from a bitstream, coefficient level information indicative of a magnitude value of the current coefficient by means of the determined Rice parameter; and obtaining the magnitude of the current coefficient by means of de-binaryzing the coefficient level information by means of the determined Rice parameter.

Systems, methods, and instrumentation for the field of video compression are disclosed herein. In an example, a video encoding device may identify a residual block. The residual block may be associated with a primary transform index and a secondary transform index. The video encoding device may determine whether the secondary transform index is non-zero and whether an enlarged non-separable primary transform (NSPT) is allowed. Based on the secondary transform being non-zero and the enlarged NSPT being allowed, the video encoder may derive a reduced residual block size from a residual block size associated with the residual block, down-sample the residual block to the reduced residual block size, apply NSPT to the reduced residual block to obtain a transformed reduced block, and transform the transformed reduced block to the residual block. Quantization, transformation, and entropy coding of the transformed reduced block are performed. The primary transform index and the secondary transform index may be encoded.

Generative Face Video Compression (“GFVC”) technology is provided to improve the performance of face video compression. The computational system is configured to: compute a relative difference metric describing the feature differences between frames; and, based on the relative difference metric, determine whether the current frame can be synthesized without entropy coding or should be re-encoded. The computational system is configured to perform two-stage training to stabilize the training of the generative adversarial network (“GAN”) in GFVC.

A method for sample adaptive offset (SAO) filtering and SAO parameter signaling in a video encoder is provided that includes determining SAO parameters for largest coding units (LCUs) of a reconstructed picture, wherein the SAO parameters include an indicator of an SAO filter type and a plurality of SAO offsets, applying SAO filtering to the reconstructed picture according to the SAO parameters, and entropy encoding LCU specific SAO information for each LCU of the reconstructed picture in an encoded video bit stream, wherein the entropy encoded LCU specific SAO information for the LCUs is interleaved with entropy encoded data for the LCUs in the encoded video bit stream. Determining SAO parameters may include determining the LCU specific SAO information to be entropy encoded for each LCU according to an SAO prediction protocol.