Abstract: Methods and apparatuses for compression of feature tensors of a neural network are provided. One or more encoding parameters for encoding the channels of a feature tensor are selected according to the importance of the channels. This enables unequal bit allocation according to the importance. Furthermore, the deployed neural network may be trained or fine-tuned considering the effect of encoding noise applied to the intermediate feature tensors. According to the present disclosure, the encoding and modified training methods are advantageous at least for employment in a collaborative intelligence framework.

Abstract: The present disclosure relates to methods and apparatuses for modifying a quantizer. In particular, within a preliminary set of quantization levels, at least one quantization level is modified based on optimization involving distortion for a predetermined set of input values. At least one another quantization level out of the preliminary set is not modified. The not modified (non-modifiable) quantization level is the minimum clipping value or the maximum clipping value. The modification may facilitate increasing the dynamic range of the quantized/inverse-quantized data. Such modified quantizer may be advantageous for employment in neural networks to compress their data such as feature maps or the like. It may improve accuracy of the neural network.

Abstract: The present disclosure relates to scalable encoding and decoding of pictures. In particular, a picture is processed by one or more network layers of a trained module to obtain base layer features. Then, enhancement layer features are obtained, e.g. by a trained network processing in sample domain. The base layer features are for use in computer vision processing. The base layer features together with enhancement layer features are for use in picture reconstruction, e.g. for human vision. The base layer features and the enhancement layer features are coded in a respective base layer bitstream and an enhancement layer bitstream. Accordingly, a scalable coding is provided which supports computer vision processing and/or picture reconstruction.

Abstract: In a decoder, a desired image is estimated by first retrieving coding modes from an encoded side information image. For each bitplane in the encoded side information image, syndrome bits or parity bits are decoded to obtain an estimated bitplane of quantized transform coefficients of the desired image. A quantization and a transform are applied to a prediction residual obtained using the coding modes, wherein the decoding uses the quantized transform coefficients of the encoded side information image, and is based on previously decoded bitplanes in a causal neighborhood. The estimated bitplanes of quantized transform coefficients of the desired image are combined to produce combined bitplanes. Then, an inverse quantization, an inverse transform and a prediction based on the coding modes are applied to the combined bitplanes to recover the estimate of the desired image.

Abstract: Blocks in pixel images are template matched to select candidate blocks and weights according to a structural similarity and a perceptual distortion of the blocks. The perceptual distortion is a function of a just-noticeable-distortion (JND). A filter outputs a prediction residual between the block and the candidate blocks. The prediction residual is transformed and quantized to produce a quantized prediction residual using the JND. The matching and quantizing is optimized jointly using the perceptual distortion. Then, the quantized prediction residual and the weights are entropy encoded into a bit-stream for later decoding.

Abstract: A communications system including a plurality of mobile user computing devices, and a service provider subsystem for enabling communications between any of the mobile user computing devices and enterprise network systems. The service provider subsystem has a plurality of clusters strategically distributed across at least one geographical region and interconnected by a global services bus. Each of the clusters includes a plurality of nodes interconnected to a distributed memory storage bus. Each of the nodes includes a service manager module for monitoring services available to the node, a service access point module for enabling communications between the node and enterprise network systems, a client access point module for enabling communications between the node and at least one of the mobile user computing devices, and a message control point module for managing communications between the client access point module and the service access point module.

Abstract: A method decodes blocks in pictures of a video in an encoded bitstream by storing previously decoded blocks in a buffer. The previously decoded blocks are displaced less than a predetermined range relative to a current block being decoded. Cached blocks are maintained in a cache. The cached blocks include a set of best matching previously decoded blocks that are displaced greater than the predetermined range relative to the current block. The bitstream is parsed to obtain a prediction indicator that determines whether the current block is predicted from the previously decoded blocks in the buffer or the cached blocks in the cache. Based on the prediction indicator, a prediction residual block is generated, and in a summation process, the prediction residual block is added to a reconstructed residual block to form a decoded block as output.

Abstract: A method for decoding a bitstream, including compressed pictures of a video, wherein each picture includes one or more slices, wherein each slice includes one or more blocks of pixels, and each pixel has a value corresponding to a color, for each slice, first obtains a reduced number of colors corresponding to the slice, wherein each color is represented as a color triplet and the reduced number of colors is less than or equal to a number of colors in the slice. Then, for each block, a prediction mode is determined, wherein an independent uniform prediction mode is included in a candidate set of prediction modes. For each block, a predictor block is generated, wherein all values of the predictor block have a uniform value according to a color index when the prediction mode is set as the independent uniform prediction mode. Lastly, the predictor block is added to a reconstructed residue block to form a decoded block as output.

Abstract: A method decodes a picture that is encoded and represented by blocks in a bitstream, by first determining, from the bitstream, motion associated with the block. Using a model, the motion is mapped to indices indicating a subset of quantized transform coefficients to be decoded from the bitstream. Then, values are assigned and reinserted to the quantized transform coefficients not in the subset.

Abstract: In a decoder, a desired image is estimated by first retrieving coding modes from an encoded side information image. For each bitplane in the encoded side information image, syndrome bits or parity bits are decoded to obtain an estimated bitplane of quantized transform coefficients of the desired image. A quantization and a transform are applied to a prediction residual obtained using the coding modes, wherein the decoding uses the quantized transform coefficients of the encoded side information image. The estimated bitplanes of quantized transform coefficients of the desired image are combined to produce combined bitplanes. Then, an inverse quantization, an inverse transform and a prediction based on the coding modes are applied to the combined bitplanes to recover the estimate of the desired image.

Abstract: In a decoder, a desired image is estimated by first retrieving coding modes from an encoded side information image. For each bitplane in the encoded side information image, syndrome bits or parity bits are decoded to obtain an estimated bitplane of quantized transform coefficients of the desired image. A quantization and a transform are applied to a prediction residual obtained using the coding modes, wherein the decoding uses the quantized transform coefficients of the encoded side information image. The estimated bitplanes of quantized transform coefficients of the desired image are combined to produce combined bitplanes. Then, an inverse quantization, an inverse transform and a prediction based on the coding modes are applied to the combined bitplanes to recover the estimate of the desired image.

Abstract: In a decoder, a desired image is estimated by first retrieving coding modes from an encoded side information image. For each bitplane in the encoded side information image, syndrome bits or parity bits are decoded to obtain an estimated bitplane of quantized transform coefficients of the desired image. A quantization and a transform are applied to a prediction residual obtained using the coding modes, wherein the decoding uses the quantized transform coefficients of the encoded side information image, and is based on previously decoded bitplanes in a causal neighborhood. The estimated bitplanes of quantized transform coefficients of the desired image are combined to produce combined bitplanes. Then, an inverse quantization, an inverse transform and a prediction based on the coding modes are applied to the combined bitplanes to recover the estimate of the desired image.

Abstract: A method codes pictures in a bitstream, wherein the bitstream includes coded pictures to obtain data for associated TUs and data for generating a transform tree, and a partitioning of coding units (CUs) into Prediction Units (PUs), and data for obtaining prediction modes or directions associated with each PU. One or more mapping tables are defined, wherein each row of each table has an associated index and a first set of transform types to be used for applying an inverse transformation to the data in TU. The first set of transform types is selected according to an index, and then a second set of transform types is applied as the inverse transformation to the data, wherein the second set of transform types is determined according to the first set of transform types and a transform-toggle flag (ttf) to obtain a reconstructed prediction residual.

Abstract: A bitstream corresponding to an encoded video is decoded. The encoded video includes a sequence of frames, and each frame is partitioned into encoded blocks. For each encoded block, an edge mode index is decoded based on an edge mode codeword and a prediction mode. The edge mode index indicates a subset of predetermined partitions selected from a partition library according to the prediction mode. The encoded block is partitioned based on the edge mode index to produce two or more block partitions. To each block partition, a coefficient rearrangement, an inverse transform and an inverse quantization is applied to produce a processed block partition. The processed block partitions are then combined into a decoded block for a video.

Abstract: A system and a method for decoding at least a portion of an image includes determining a current prediction mode based on a combination of a prediction mode residue and a function of at least one previous prediction mode and decoding the portion of the image using the current prediction mode.

Abstract: Specifically, a method decodes a picture. The picture is encoded and represented by blocks in a bitstream. For each block, a remap flag is obtained from the bit-stream. The block is either a remapped reconstructed block or a non-remapped reconstructed block. Either the non-mapped reconstructed block or an inverse remapped reconstructed block is output according to the remap flag. The remapped reconstructed block maximizes a similarity with the neighboring blocks, as compared to the similarity of the non-mapped reconstructed block and the neighboring blocks, by applying point operations to the remapped reconstructed block.

Abstract: A video encoded as a bit stream is decoded by maintaining a set of dictionaries generated from decoded prediction residual signals, wherein elements of the set of dictionaries have associated indices. A current macroblock is entropy decoded and inverse quantized to produce decoded coefficients. For the current macroblock, a particular dictionary of the set of dictionaries is selected according to a prediction mode signaled in the bit stream, and particular elements of the particular dictionary are selected according to a copy mode signal in the bit stream and the associated index. The particular elements is scaled and combined, using the decoded coefficients, to reconstruct a current decoded macroblock prediction residual signal. Then, the current decoded macroblock prediction residual signal is combined with previously decoded macroblocks to generate an output macroblock of a reconstructed video, wherein the steps are performed in a decoder.

Abstract: Blocks in pixel images are template matched to select candidate blocks and weights according to a structural similarity and a perceptual distortion of the blocks. The perceptual distortion is a function of a just-noticeable-distortion (JND). A filter outputs a prediction residual between the block and the candidate blocks. The prediction residual is transformed and quantized to produce a quantized prediction residual using the JND. The matching and quantizing is optimized jointly using the perceptual distortion. Then, the quantized prediction residual and the weights are entropy encoded into a bit-stream for later decoding.

Abstract: A bitstream includes coded pictures, and split-flags for generating a transform tree. The bit stream is a partitioning of coding units (CUs) into Prediction Units (PUs). The transform tree is generated according to the split-flags. Nodes in the transform tree represent transform units (TU) associated with the CUs. The generation splits each TU only if the corresponding split-flag is set. For each PU that includes multiple TUs, the multiple TUs are merged into a larger TU, and the transform tree is modified according to the splitting and merging. Then, data contained in each PU can be decoded using the TVs associated with the PU according to the transform tree.

Abstract: A communications system including a plurality of mobile user computing devices, and a service provider subsystem for enabling communications between any of the mobile user computing devices and enterprise network systems. The service provider subsystem has a plurality of clusters strategically distributed across at least one geographical region and interconnected by a global services bus. Each of the clusters includes a plurality of nodes interconnected to a distributed memory storage bus. Each of the nodes includes a service manager module for monitoring services available to the node, a service access point module for enabling communications between the node and enterprise network systems, a client access point module for enabling communications between the node and at least one of the mobile user computing devices, and a message control point module for managing communications between the client access point module and the service access point module.