Interdigital Ce Patent Holdings, Sas

Interdigital Vc Holdings, Inc.

Interdigital Madison Patent Holdings, Sas

Interdigital Madison Holdings, Sas

Fabrice LeLeannec

Franck Galpin

Tangi Poirier

Edouard Francois

Philippe Bordes

Fabrice Urban

Gagan Bihari Rath

Philippe De Lagrange

Karam Naser

Antoine Robert

Pierre Andrivon

Ya Chen

Fabien Racape

Michel Kerdranvat

Thierry Dumas

Gaëlle Martin-Cocher

Baptiste Esteban

Asymmetric binary trees are used in combination with coding tools such as transform unit tiling, affine motion compensation, decoder side motion vector refinement, and bidirectional optical flow. In another embodiment asymmetric binary trees is used in combination with subblock temporal motion vector prediction. Embodiments enable tiling of coding blocks into subblocks in accordance with coding tools such as used in Versatile Video Coding.

Combining ABT with VVC sub-block-based coding tools

In an encoding apparatus or a decoding apparatus, an encoding method or decoding method combines various coding tools with intra sub partitioning mode in which a coding unit is split in multiple sub partitions. At least one embodiment particularly aims at the selection of primary transforms among a set of multiple transforms for video encoding or decoding of intra sub block partitions. At least one embodiment particularly aims at intra coding using Matrix-Based Intra Prediction mode combined with Intra Sub Partition mode. At least one embodiment particularly aims at intra coding using Multi-Reference-Line intra prediction mode combined with Intra Sub Partition mode.

Intra sub partitions for video encoding and decoding combined with multiple transform selection, matrix weighted intra prediction or multi-reference-line intra prediction

Edges of the plurality of reference blocks are adjacent to a current block of a picture. A single transform is applied to samples of each reference block. A single transform is applied to samples of the current block. A decoding method includes decoding a plurality of reference blocks, obtaining a plurality of filtered reference blocks by filtering with a deblocking filter along the edges of the plurality of decoded reference blocks, and decoding the current block by intra prediction from the plurality of filtered reference blocks. An encoding method includes encoding and reconstructing a plurality of reference blocks, obtaining a plurality of filtered reference blocks by filtering with a deblocking filter along the edges of the plurality of reconstructed reference blocks, and encoding the current block by intra prediction from the plurality of filtered reference blocks.

Method and a device for decoding an intra predicted block of a picture and corresponding coding method and coding device

Various implementations for video encoding and decoding are presented involving determining, for a block being encoded or decoded in a picture, a coefficient group mode for coding at least one coefficient of a set of transform coefficients of the image block; encoding or decoding the set of transform coefficients of the image block responsive to the coefficient group mode. The coefficient group mode can be determined from at least one of image block size, number of non-zero transform coefficients inside the image block, position of a transform coefficient inside the image block, decoded syntax element, wherein the coefficient group mode can specify whether a coefficient group significance flag is coded/decoded, indicating that at least one coefficient is non-zero inside a coefficient group and/or specify at least one size of the coefficient group.

Method and apparatus for video encoding and decoding based on adaptive coefficient group

A method and apparatus to improve compression efficiency in a video compression scheme enables separate control of a local chroma quantization parameter. The control can be used to disable a local chroma quantization parameter. One embodiment uses an existing local luma quantization parameter control flag for controlling local chroma quantization parameter. Another embodiment includes a separate local chroma quantization parameter control flag. Bitstream syntax is used to convey the flags.

High-level constraint flag for local chroma quantization parameter control

A video coding system is provided that performs intra prediction in a mode using a neural network for block of only a set of specific block sizes. The signaling of this mode is designed to be efficient in terms of rate-distortion under this constraint. Different transformations of the context of a block and the neural network prediction of this block are introduced in order to use one single neural network for predicting blocks of several sizes, as well as the corresponding signaling. The neural network-based prediction mode considers both luminance blocks and chrominance blocks. The video coding system comprises encoder and decoder apparatuses, encoding, decoding and signal generation methods and a signal carrying information corresponding to the described coding mode.

Neural network-based intra prediction for video encoding or decoding

A method and apparatus for encoding/decoding a picture use a secondary transform for video encoding and decoding in INTRA mode and more particularly use a secondary transform for the transform units of a coding unit in intra sub partition mode. Different signaling and selection variants are proposed according to various parameters of the coding unit and transform unit.

Secondary transform for video encoding and decoding

Coding mode information of a current block is propagated to neighbor blocks even when the coding mode is not selected for a current block. Such propagation may be limited, for example to one block only or to multiple generation of copies. Different propagation modes are proposed: left to right direction, top to bottom direction, bottom-right diagonal direction, first available information, or last coded information. The coding mode propagation allows to improve further predictions since a further block will benefit from the information from its neighbors.

Coding mode information propagation for video coding

Video coding tools can be controlled by including syntax in a video bitstream that makes better use of video decoding resources. An encoder inserts syntax into a video bitstream to enable a decoder to parse the bitstream and easily control which tools combinations are enabled, which combinations are not permitted, and which tools are activated for various components in a multiple component bitstream, leading to potential parallelization of bitstream decoding.

Managing coding tools combinations and restrictions

A video encoder or decoder processes portions of video using intra coding and with variable block shapes. Under the general aspects, intra prediction directions are adapted to the block shape or aspect ratio. The general aspects described adaptively replace several conventional angular intra prediction modes with wide-angle intra prediction modes for non-square blocks. The total number of angular directions for any block can remain the same as in prior video coding standards or can be reduced or expanded. To keep the same number of prediction directions, directions along a shorter rectangular edge of a target block are removed but replaced with addition directions along the longer edge of the rectangular target block. For directions with wide angles, reference sample smoothing is performed. Signaling can be provided to inform a corresponding decoder regarding the prediction modes and directions.

Block shape adaptive intra prediction directions for quadtree-binary tree

A method and apparatus to improve compression efficiency in a video compression scheme enables use of new tools with multiple transform selection. In one embodiment, transform pair selection is based on a flag indicative of low-frequency non-separable transforms. In another embodiment, transform pair selection is based on a flag indicative of low-frequency non-separable transforms and on a flag indicative of matrix-based intra prediction. In another embodiment, when an implicit multiple transform selection mode is used, transform pair selection is based on a flag indicative of low-frequency non-separable transforms. Bitstream syntax is used to convey the flags.

Transform selection for implicit multiple transform selection

Different implementations are described, particularly implementations for video encoding and decoding are presented. Accordingly, the encoding or decoding comprises obtaining a merging candidate index that enables deriving both a first merging candidate index and a second merging candidate index in a triangular shape based motion compensation candidates list in list in use in a triangular shape based motion compensation. According to a particular characteristic, a mapping function determines the derivation process of both indices among a subset of possible combinations of merge candidate indices.

Signaling of merge indices for triangle partitions

To encode a picture, a coding tree unit (CTU) in the picture is partitioned by a quadtree structure, and the quadtree leaf nodes can be further partitioned by a multi-type tree (MTT) structure. To increase the set of reachable coding tree nodes and leaves, we propose to increase the maximum allowed MTT hierarchy depth to be twice the difference between the CTU size and the minimum allowed size for a CU. The maximum allowed MTT hierarchy depth can be specified for all QT levels in order to provide more flexibility in the split tree. Alternatively, only two levels of maximum allowed MTT depth are signaled: one when QT splits are allowed, and another one when no more QT splits are allowed. In addition, an upper bound can be set for the minimum allowed coding block size, based on the coding tree unit size or the maximum allowed transform size.

Coding tree partitioning

Video content frequently seen in gaming content and having a higher probability of being coded in transform skip mode is coded separately from remaining portions of a video coding block before being combined with regular coded portions. This enables more efficient coding of video by reducing the switching between regular coding modes and transform skip modes. The regular coded portion and the transform skip coded portion are combined in an encoder. At a corresponding decoder, inverse operations occur, such that the transform skipped residual portions are decoded and the regularly coded portions are decoded before the portions are combined for a resultant decoded video block.

Hybrid texture particle coding mode improvements

To encode a coding unit in the inter modes, there exist multiple inter modes that signal the motion information using a Motion Vector Predictor (MW) list. The MW lists for different coding modes include spatial predictors. In one embodiment, for all coding modes that use MVP lists, the top spatial predictor is first added to the motion vector predictor list before the left spatial predictor. In another embodiment, a subset of coding modes adds the top spatial predictor before the left spatial predictor, and another subset of coding modes adds the left spatial predictor before the top spatial predictor. For example, all merge modes add the top spatial predictor first, and all AMVP modes add the left spatial predictor first. In another example, all non-sub-block modes add the top spatial predictor first, and all sub-block modes add the left spatial predictor first.

Growing community of inventors

BETTON, France

Fabrice LeLeannec