Sharp Kabushiki Kaisha Corporation

Sharp Laboratories of America, Inc

Dolby International Ab

Velos Media, LLC

Fg Innovation Company Limited

Huawei Technologies Co., Limited

Amazon Technologies, Inc.

Other

Huawei Technology Co., Ltd.

Dobly International Ab

Velos Media International Limited

Christopher A Segall

Kiran Mukesh Misra

Jie Zhao

Sachin G Deshpande

Louis Joseph Kerofsky

Seung-Hwan Kim

Frank Bossen

Yeping Su

Weijia Zhu

Philip Cowan

Jonatan Samuelsson

Anustup Kumar Choudhury

Tomohiro Ikai

Byeongdoo Choi

Sheau Ng

Anustup Kumar A Choudhury

Michael Horowitz

Petrus J L Van Beek

Zhan Ma

Shawmin Lei

Hae-Jong Seo

Xu Chen

Wei-Jia Zhu

Mohammad Aghagolzadeh

Shijun Sun

M Ibrahim Sezan

Shaw-min Lei

Gary Alan Feather

Jon Arthur Fairhurst

Junlan Yang

Tianying Ji

Byung Tae Oh

Mark DeNies

Ron Rubinstein

Sachi N G Deshpande

The present disclosure relates to methods, apparatus, systems, and non-transitory computer-readable storage media for training and using a multi-scale machine learning model for the enhancement of compressed video. According to some examples, a computer-implemented method includes receiving a video at a content delivery service; performing an encode on a frame of the video by the content delivery service that converts the frame from a pixel domain to a transform domain and back to the pixel domain to generate first pixel values and a first residual for a block of the frame at a first resolution; generating a first set of features, by a machine learning model of the content delivery service, for an input, at a first resolution, of the first pixel values and the first residual of the block; generating a second set of features, by the machine learning model of the content delivery service, for an input, at a second lower resolution, of second pixel values and a second residual of the block; upsampling the second set of features to the first resolution to generate an upsampled second set of features; generating a modified version of the frame based on the first set of features and the upsampled second set of features; and transmitting the modified version of the frame to a frame buffer or from the content delivery service to a viewer device.

Computer-implemented multi-scale machine learning model for the enhancement of compressed video

This disclosure relates to video coding and more particularly to techniques for performing deblocking of reconstructed video data. According to an aspect of an invention, reconstructed video data having a 4:2:2 chroma format is received and for a four line decision segment corresponding to a vertical deblocking boundary for a chroma component of the video data, a gradient value is computed using samples included in the first and fourth lines.

Systems and methods for applying deblocking filters to reconstructed video data

A method of filtering reconstructed video data is disclosed. The method comprising: inputting reconstructed luma component sample values; deriving a filtered sample value by using cross component filter coefficients and the reconstructed luma component sample values prior to an adaptive loop filtering process; deriving a refinement value for chroma components by using the filtered sample value; and deriving a refined chroma sample value by using a sum of a sample value of chroma components and the refinement value for chroma components.

Systems and methods for reducing a reconstruction error in video coding based on a cross-component correlation

Systems and methods for applying deblocking filters to reconstructed video data are disclosed. Sample values in adjacent reconstructed video blocks are modified according to multiple passes of a deblocking filter. A filtering pass may correspond to processing or constructing of all or subset of samples to be deblocked. The number of processing or constructing for each sample in a given pass may correspond to the pass index or order.

The present disclosure relates to methods, apparatus, systems, and non-transitory computer-readable storage media for video coding using an implicit video frame output process. According to some examples, a computer-implemented method includes receiving a video at a content delivery service; encoding, by the content delivery service, the video into an encoded video; generating, by the content delivery service, at least one open bitstream unit from the encoded video according to a video coding format that does not utilize a show existing frame syntax element set to one to indicate a frame in a reference picture buffer of a decoder is to be displayed; and transmitting the at least one open bitstream unit from the content delivery service to the decoder.

Computer-implemented method and apparatus for video coding using an implicit video frame output process

The present disclosure relates to methods, apparatus, systems, and non-transitory computer-readable storage media for video coding using super-resolution restoration with residual frame coding. According to some examples, a computer-implemented method includes receiving a coded frame of a video; performing a video coding on the coded frame of the video to generate a resultant for the coded frame at a second lower resolution than a first resolution; upsampling the resultant in at least a vertical direction to a higher resolution than the second lower resolution to generate an upsampled resultant; generating a decoded frame based on at least the upsampled resultant; and transmitting the decoded frame to a frame buffer or to a display device.

Computer-implemented method and apparatus for video coding using super-resolution restoration with residual frame coding

Technique for initialization of encoders and decoders. In some cases, the decoder receives a slice and identifies if the slice is either a forward predicted B-slice or a backward predicted B-slice, and not both a forward and backward predicted B-slice, and based upon this identification initializes, using a P-slice technique, a context associated with the slice.

Context initialization based on slice header flag and slice type

Methods and systems for improving coding decoding efficiency of video by providing a syntax modeler, a buffer, and a decoder. The syntax modeler may associate a first sequence of symbols with syntax elements. The buffer may store tables, each represented by a symbol in the first sequence, and each used to associate a respective symbol in a second sequence of symbols with encoded data. The decoder decodes the data into a bitstream using the second sequence retrieved from a table.

Nested entropy encoding

A method for tracking a reference picture on an electronic device is described. The method includes receiving a bitstream. The method also includes decoding a portion of the bitstream to produce a decoded reference picture. The method further includes tracking the decoded reference picture in a decoded picture buffer (DPB) with reduced overhead referencing. The method additionally includes decoding a picture based on the decoded reference picture.

Tracking a reference picture on an electronic device

A method of encoding video data by a device is provided. Transform coefficient values associated with a chroma coding block are received. The chroma coding block is one of chroma coding blocks resulting from partitioning a chroma coding tree block and misaligned with a luma coding tree block corresponding to the chroma coding tree block and partitioned into luma coding blocks which are organized into quantization groups such that luma quantization parameters are determined for the luma coding tree block. One of the luma quantization parameters that is mapped to the chroma coding block is identified. A predictive quantization parameter is derived by using a value derived from the identified luma quantization parameter. A chroma quantization parameter is determined by adding a delta quantization parameter associated with each of the chroma coding blocks to the predictive quantization parameter. Level values are generated based on the determined chroma quantization parameter.

Systems and methods for varying quantization parameters

A method of filtering reconstructed video data is provided. The method receives an array of reconstructed luma samples including a P-block and a Q-block that are both adjacent to an intra sub-partitions (ISP) edge, and determines a first maximum filter length based on whether a first size of an edge of one of the P-block and the Q-block is less than 32. The edge of the one of the P-block and the Q-block is perpendicular to the ISP edge. The method further deblocks the array of the reconstructed luma samples based on the determined first maximum filter length.

This disclosure relates to video coding and more particularly to techniques for signaling picture output resolution for coded video. According to an aspect of an invention, a first flag syntax element indicating whether conformance cropping window offset parameters are present in a sequence parameter set is parsed, wherein the conformance cropping window offset parameters specify a cropping window that is applied to pictures with a width equal to a maximum width and a height equal to a maximum height.

Systems and methods for signaling picture output resolution in video coding

A method of filtering reconstructed video data is disclosed. The method comprising: parsing a first syntax element used for setting cross-component filter coefficients; inputting a reconstructed luma picture sample array; deriving luma locations by using location corresponding to a current chroma sample; deriving a filter coefficient array by using the cross-component filter coefficients; deriving a variable by using the filter coefficient array and the reconstructed luma picture sample array defined by the luma locations; and deriving a scaled variable by using the variable, wherein the variable is modified by a sum of a sample of a current chroma block, which is defined by a predetermined location, and the scaled variable.

A method for decoding video includes receiving quantized coefficients representative of a block of video representative of a plurality of pixels. The quantized coefficients are dequantized based upon a function of a remainder. The dequantized coefficients are inverse transformed to determine a decoded residue.

Video decoder with reduced dynamic range transform with inverse transform shifting memory

A decoding device for filtering reconstructed video data is provided. The decoding device includes computer-readable media storing instructions that, when executed, cause the decoding device to: receive an edge type variable specifying whether a vertical edge or a horizontal edge is being filtered in a current block of the reconstructed video data; derive a maximum filter length of a deblocking filter based on the edge type variable and one or more conditions relating to at least one of a size of a transform block associated with the current block or a size of an adjacent transform block; and perform a deblocking filtering, by the deblocking filter, on the current block of the reconstructed video data based on the maximum filter length, wherein a value of a luma filtered sample of the reconstructed video data is clipped based on a position of the luma filtered sample of the reconstructed video data.

Growing community of inventors

Camas, WA, United States of America

Christopher A Segall