IP Library › Granted Patent US 12,231,643
Granted Patent B2
US 12,231,643 · App. 17/627,377 · Granted Feb 18, 2025

Method and apparatus for video encoding and decoding with matrix based intra-prediction

Inventors: Franck Galpin (Thorigne-Fouillard, FR); Fabien Racape (San Francisco, CA); Jean Begaint (Menlo Park, CA); Swayambhoo Jain (Milpitas, CA); Shahab Hamidi-Rad (Sunnyvale, CA)
Assignee: InterDigital Madison Patent Holdings, SAS
H04N19/132H04N19/105H04N19/159H04N19/176H04N19/70
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12,231,643
App. No.
17/627,377
Granted
Feb 18, 2025
Kind
B2
Abstract

Different implementations are described, particularly implementations for video encoding and decoding based on linear weighted intra prediction, also called matrix based intra prediction, are presented. Accordingly, for a block being encoded or decoded in linear weighted intra prediction, obtaining intra predicted samples from at least two matrix-vector products between at least two selected weight matrices of reduced size and a set of neighboring reference samples. Advantageously, such arrangement allows to reduce the amount of memory for storing data and to reduce the complexity of the intra prediction samples computation.

Claims (42)

1. A method comprising:

encoding a block of a picture of a video in intra prediction mode using linear weighted prediction,

wherein the encoding comprises:

determining a set of n neighboring reference samples x;

selecting at least a first weight matrix G of size m×r, a second weight matrix H T of size r×n, and an associated bias vector b among a set of weight matrices and associated bias vectors based on a mode of linear weighted intra prediction and a block shape, the first weight matrix G having a size m×r and the second weight matrix H T having a size r×n, where m, n, r are integers with r smaller than m and r smaller than n; and

obtaining intra predicted samples ŷ of the block from a first product of matrices H T x, a second product of matrices G(H T x), and adding associated bias b resulting in ŷ=G(H T x)+b.

2. The method of claim 1 , wherein the first and second weight matrices G and H T are obtained from low rank processing of a weight matrix M of size m×n and wherein an estimate M of the weight matrix M is the product of the first and second weight matrices G and H T , {circumflex over (M)}=GH T .

3. The method of claim 1 , wherein r is determined to reduce matrices product complexity.

4. The method of claim 1 , wherein r is determined responsive to Single Value Decomposition SVD of the weight matrix M.

5. The method of claim 1 , wherein, for a block of size width*height, the set of neighboring reference samples x comprises all reconstructed samples from a neighboring top line of size width and from a neighboring left line of size height.

6. The method of claim 1 , wherein for a block of size width*height, the set of neighboring reference samples x comprises all reconstructed samples from a neighboring top line of size 2*width and from a neighboring left line of size 2*height.

7. The method of claim 1 , wherein at least one high level syntax element representative of the set of weight matrix and associated bias are signaled in a Picture Parameter Set (PPS) so that all blocks in a frame using linear weighted prediction uses the signaled set of weight matrices.

8. The method of claim 1 , wherein at least one high level syntax element representative of the set of weight matrix and associated bias are signaled in a Sequence Parameter Set (SPS) so that all blocks in a sequence using linear weighted prediction uses the signaled set of weight matrices.

9. The method of claim 1 , wherein a bit depth of the first product of matrices and of the second product of matrices is increased with regard to a bit depth of intra predicted samples.

10. A method comprising:

decoding a block of a picture of a video, the block being coded in intra prediction mode using linear weighted prediction,

wherein the decoding comprises:

determining a set of n neighboring reference samples x;

selecting at least a first weight matrix G of size m×r, a second weight matrix H T of size r×n, and an associated bias vector b among a set of weight matrices and associated bias vectors based on a mode of linear weighted intra prediction and a block shape, the first weight matrix G having a size m×r and the second weight matrix H T having a size r×n, where m, n, r are integers with r smaller than m and r smaller than n; and

obtaining intra predicted samples ŷ of the block from a first product of matrices H T x, a second product of matrices G(H T x), and adding associated bias b resulting in ŷ=G (H T x)+b.

11. The method of claim 10 , wherein the first and second weight matrices G and H T are obtained from low rank processing of a weight matrix M of size m×n, and wherein an estimate M of the weight matrix M is the product of the first and second weight matrices G and H T , {circumflex over (M)}=GH T .

12. The method of claim 10 , wherein r is determined to reduce matrices product complexity.

13. The method of claim 10 , wherein r is determined responsive to Single Value Decomposition SVD of the weight matrix M.

14. The method of claim 10 , wherein, for a block of size width*height, the set of neighboring reference samples x comprises all reconstructed samples from a neighboring top line of size width and from a neighboring left line of size height.

15. The method of claim 10 , wherein for a block of size width*height, the set of neighboring reference samples x comprises all reconstructed samples from a neighboring top line of size 2*width and from a neighboring left line of size 2*height.

16. The method of claim 10 , wherein at least one high level syntax element representative of the set of weight matrix and associated bias are signaled in a Picture Parameter Set (PPS) so that all blocks in a frame using linear weighted prediction uses the signaled set of weight matrices.

17. The method of claim 10 , wherein at least one high level syntax element representative of the set of weight matrix and associated bias are signaled in a Sequence Parameter Set (SPS) so that all blocks in a sequence using linear weighted prediction uses the signaled set of weight matrices.

18. The method of claim 10 , wherein a bit depth of the first product of matrices and of the second product of matrices is increased with regard to a bit depth of intra predicted samples.

19. An apparatus comprising one or more processors, and at least one memory, the one or more processors configured for:

encoding a block of a picture of a video in intra prediction mode using linear weighted prediction,

wherein intra predicted samples of the block are obtained by:

determining a set of n neighboring reference samples x;

selecting at least a first weight matrix G of size m×r, a second weight matrix H T of size r×n, and an associated bias vector b among a set of weight matrices and associated bias vectors based on a mode of linear weighted intra prediction and a block shape, the first weight matrix G having a size m×r and the second weight matrix H T having a size r×n, where m, n, r are integers with r smaller than m and r smaller than n; and

obtaining intra predicted samples ŷ of the block from a first product of matrices H T x, a second product of matrices G (H T x), and adding associated bias b resulting in ŷ=G (H T x)+b.

20. An apparatus comprising one or more processors, and at least one memory, the one or more processors configured for:

decoding a block of a picture of a video, the block being coded in intra prediction mode using linear weighted prediction,

wherein intra predicted samples of the block are obtained by:

determining a set of n neighboring reference samples x;

selecting at least a first weight matrix G of size m×r, a second weight matrix H T of size r×n, and an associated bias vector b among a set of weight matrices and associated bias vectors based on a mode of linear weighted intra prediction and a block shape, the first weight matrix G having a size m×r and the second weight matrix H T having a size r×n, where m, n, r are integers with r smaller than m and r smaller than n; and

obtaining intra predicted samples ŷ of the block from a first product of matrices H T x, a second product of matrices G (H T x), and adding associated bias b resulting in ŷ=G (H T x)+b.

21. A non-transitory computer readable medium having stored instructions which, when executed by one or more processors, cause the one or more processors to perform the method of claim 1 .

22. A non-transitory computer readable medium having stored instructions which, when executed by one or more processors, cause the one or more processors to perform the method of claim 10 .

Assignments (5)
CORRECTIVE ASSIGNMENT TO CORRECT THE 5TH ASSIGNOR'S NAME PREVIOUSLY RECORDED ON REEL 063240 FRAME 0487. ASSIGNOR(S) HEREBY CONFIRMS THE ASSIGNMENT. Recorded Aug 15, 2023
From: GALPIN, FRANCK; RACAPE, FABIEN; BEGAINT, JEAN; JAIN, SWAYAMBHOO; HAMIDI-RAD, SHAHAB
To: INTERDIGITAL VC HOLDINGS, INC.
Reel/Frame 064588/0190 →
CORRECTIVE ASSIGNMENT TO CORRECT THE 5TH ASSIGNOR'S NAME AND ASSIGNEE ADDRESS PREVIOUSLY RECORDED ON REEL 061629 FRAME 0019. ASSIGNOR(S) HEREBY CONFIRMS THE ASSIGNMENT. Recorded Aug 15, 2023
From: RACAPE, FABIEN; GALPIN, FRANCK; BEGAINT, JEAN; JAIN, SWAYAMBHOO; HAMIDI-RAD, SHAHAB
To: INTERDIGITAL VC HOLDINGS, INC.
Reel/Frame 064646/0390 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Apr 3, 2023
From: GALPIN, FRANCK; RACAPE, FABIEN; BEGAINT, JEAN; JAIN, SWAYAMBHOO; HAMIDI-RAB, SHAHAB
To: INTERDIGITAL VC HOLDINGS, INC.
Reel/Frame 063240/0487 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Jan 4, 2023
From: INTERDIGITAL VC HOLDINGS, INC.
To: INTERDIGITAL MADISON PATENT HOLDINGS, SAS
Reel/Frame 062291/0394 →
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Nov 2, 2022
From: RACAPE, FABIEN; GALPIN, FRANCK; BEGAINT, JEAN; JAIN, SWAYAMBHOO; HAMIDI-RAB, SHAHAB
To: INTERDIGITAL VC HOLDINGS, INC.
Reel/Frame 061629/0019 →
Priority Claims (1)
EP 19305969 · Jul 22, 2019 · regional
Continuity (1)
Related Publication 20220256161A1 · Aug 11, 2022
References Cited (22)
US 20140252857A1 · Wu et al. · 2014 [cited by applicant]
US 20140269939A1 · Guo et al. · 2014 [cited by applicant]
US 20160087834A1 · Zhao et al. · 2016 [cited by applicant]
US 20170094285A1 · Said et al. · 2017 [cited by applicant]
CN 108141608A · 2018 [cited by applicant]
CN 108988988A · 2018 [cited by applicant]
EP 3484151A1 · 2019 [cited by applicant]
WO 2018219925A1 · 2018 [cited by applicant]
Jonathan et al. (“CE3: Affine linear weighted intra prediction (CE3-4.1, CE3-4.2)”, Joint Video Experts Team (JVET) of ITU-T SG 16 WP 3 and ISO/IEC JTC 1/SC 29/WG 11, Published 14th Meeting: Geneva, CH, Mar. 19-27, 2019… [cited by examiner]
Fabien et al. (“Report on CE2 results from Interdigital”, International Organisation for Standardisation Organisation Internationale De Normalisation ISO/IEC JTC1/SC29/WG11 Coding of Moving Pictures and Audio, published… [cited by examiner]
Benjamin et al. (“Versatile Video Coding (Draft 5)”, Joint Video Experts Team (JVET) of ITU-T SG 16 WP 3 and ISO/IEC JTC 1/SC 29/WG 11, published on 14th Meeting: Geneva, CH, Mar. 19-27, 2019) (Year: 2019). [cited by examiner]
Bross et al., “Versatile Video Coding (Draft 5)”, Joint Video Experts Team (JVET) of ITU-T SG 16 WP 3 and ISO/IEC JTC 1/SC 29/WG 11, Document: JVET-N1001-v8, 14th Meeting, Geneva, Switzerland, Mar. 19, 2019, 397 pages. [cited by applicant]
Pan, et al., “Inversion of Displacement Operators”, Society for Industrial and Applied Mathematics (SIAM), Journal on Matrix Analysis and Applications, vol. 24, No. 3, 2003, 18 pages. [cited by applicant]
Bross et al., “Versatile Video Coding (Draft 5)”, Joint Video Experts Team (JVET) of ITU-T SG 16 WP 3 and ISO/IEC JTC 1/SC 29/WG 11, Document: JVET-N1001-v10, 14th Meeting, Geneva, Switzerland, Mar. 19, 2019, 84 pages. [cited by applicant]
Racape et al., “Report on CE2 results from Interdigital”, International Organization for Standardization, ISO/IEC JTC1/SC29/WG11—Coding of Moving Pictures and Audio, Document: MPEG2019/m49258, Gothenburg, Sweden, Jul. 2… [cited by applicant]
Anonymous, “Reference software for ITU-T H.265 high efficiency video coding”, International Telecommunication Union, Series H: Audiovisual and Multimedia Systems Infrastructure of audiovisual services—Coding of moving v… [cited by applicant]
Wang et al., “A Low-rank Matrix Completion Based Intra Prediction for H.264/AVC”, 2011 IEEE 13th International Workshop on Multimedia Signal Processing, Oct. 17, 2011, Hangzhou, China, 6 pages. [cited by applicant]
Albrecht et al., “Description of SDR, HDR, and 360° video coding technology proposal by Fraunhofer HHI”, Joint Video Experts Team (JVET) of ITU-T SG 16 WP 3 and ISO/IEC JTC 1/SC 29/WG 11, Document: JVET-J0014-v4, 10th M… [cited by applicant]
Anonymous, “Transmission of Non-Telephone Signals: Information Technology—Generic Coding of Moving Pictures and Associated Audio Information: Systems”, International Telecommunication Union, ITU-T Telecommunication Stan… [cited by applicant]
Anonymous, “Series H: Audiovisual and Multimedia Systems, Infrastructure of audiovisual services—Coding of moving video—Information Technology—Generic coding of moving pictures and associated audio information: Video”. … [cited by applicant]
Pfaff et al., “CE3: Affine linear weighted intra prediction (CE3-4.1, CE3-4.2)”, Joint Video Exploration Team (JVET) of ITU-T SG 16 WP 3 and ISO/IEC JTC 1/SC 29/WG 11, Document: JVET-N0217, 14th Meeting, Geneva, Switzer… [cited by applicant]
Kondo et al., “Non-CE3: On rounding shift of MIP”, Joint Video Experts Team (JVET) of ITU-T SG 16 WP 3 and ISO/IEC JTC 1/SC 29/WG 11, Document: JVET-O0408-v2, 15th Meeting, Gothenburg, Sweden, Jul. 3, 2019, 4 pages. [cited by applicant]