IP Library › Granted Patent US 12,581,081
Granted Patent B2
US 12,581,081 · App. 18/638,301 · Granted Mar 17, 2026

Affine mode signaling in video encoding and decoding

Inventors: Franck Galpin (Thorigne-Fouillard, FR); Fabrice Le Leannec (Betton, FR); Philippe Bordes (Laille, FR)
Assignee: InterDigital VC Holdings, Inc.
H04N19/13H04N19/105H04N19/159H04N19/172H04N19/46
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 12,581,081
App. No.
18/638,301
Granted
Mar 17, 2026
Kind
B2
Abstract

In general, encoding or decoding a picture part can involve a first CABAC probability model associated with a first flag indicating use of an affine mode and a second CABAC probability model associated with a second flag indicating use of either the affine mode or a second mode different from the affine mode, where the first and second CABAC probability models are different and are determined independently.

Claims (69)

1 . A method, comprising:

deriving a CABAC context index associated with a first CABAC context, the first CABAC context being associated with a first flag signaled at a block level of a video, the first flag indicating use of an affine motion model for the block when the block is in an AMVP mode;

the CABAC context index being associated with a second CABAC context, the second CABAC context being associated with a second flag signaled at the block level;

the second flag indicating use of an affine motion model for the block when the block is in a merge mode;

wherein the CABAC context index is derived using a same context derivation in the AMVP mode and in the merge mode;

the AMVP mode being a prediction mode wherein a motion vector difference with a motion predictor is coded for the block, and

the merge mode being a prediction mode wherein the motion vector difference is inferred to be zero for the block;

obtaining a first CABAC probability model associated with the first CABAC context, when the block is in the AMVP mode;

obtaining a second CABAC probability model associated with the second CABAC context, when the block is in the merge mode, the second CABAC probability model being different from the first CABAC probability model;

decoding the first flag based on the first CABAC context and the first CABAC probability model when the block is in the AMVP mode and decoding the second flag based on the second CABAC context and the second CABAC probability model when the block is in the merge mode; and

decoding the block based on the first flag when the block is in the AMVP mode and decoding the block based the second flag when the block is in the merge mode.

2 . The method of claim 1 , wherein the merge mode comprises one of merge, SbTMVP, mmvd, or DMVR.

3 . A non-transitory computer readable medium containing computer program comprising instructions for performing the method of claim 1 when executed by one or more processors.

4 . An apparatus, comprising:

one or more processors, wherein the one or more processors are configured for:

deriving a CABAC context index associated with a first CABAC context, the first CABAC context being associated with a first flag signaled at a block level of a video, the first flag indicating use of an affine motion model for the block when the block is in an AMVP mode;

the CABAC context index being associated with a second CABAC context, the second CABAC context being associated with a second flag signaled at the block level;

the second flag indicating use of an affine motion model for the block when the block is in a merge mode;

wherein the CABAC context index is derived using a same context derivation in the AMVP mode and in the merge mode;

the AMVP mode being a prediction mode wherein a motion vector difference with a motion predictor is coded for the block, and

the merge mode being a prediction mode wherein the motion vector difference is inferred to be zero for the block;

obtaining a first CABAC probability model associated with the first CABAC context, when the block is in the AMVP mode;

obtaining a second CABAC probability model associated with the second CABAC context, when the block is in the merge mode, the second CABAC probability model being different from the first CABAC probability model;

decoding the first flag based on the first CABAC context and the first CABAC probability model when the block is in the AMVP mode and decoding the second flag based on the second CABAC context and the second CABAC probability model when the block is in the merge mode; and

decoding the block based on the first flag when the block is in the AMVP mode and decoding the block based the second flag when the block is in the merge mode.

5 . The apparatus of claim 4 , wherein the merge mode comprises one of SbTMVP, mmvd, or DMVR.

6 . The apparatus of claim 4 , further comprising at least one of:

an antenna configured to receive a signal, the signal including data representative of the block;

a band limiter configured to limit the received signal to a band of frequencies that includes the data representative of the block; or

a display configured to display an image from the block.

7 . A method, comprising:

deriving a CABAC context index associated with a first CABAC context, the first CABAC context being associated with a first flag signaled at a block level of a video, the first flag indicating use of an affine motion model for the block when the block is in an AMVP mode;

the CABAC context index being associated with a second CABAC context, the second CABAC context being associated with a second flag signaled at the block level;

the second flag indicating use of an affine motion model for the block when the block is in a merge mode;

wherein the CABAC context index is derived using a same context derivation in the AMVP mode and in the merge mode;

the AMVP mode being a prediction mode wherein a motion vector difference with a motion predictor is coded for the block, and

the merge mode being a prediction mode wherein the motion vector difference is inferred to be zero for the block;

obtaining a first CABAC probability model associated with the first CABAC context, when the block is in the AMVP mode;

obtaining a second CABAC probability model associated with the second CABAC context, when the block is in the merge mode, the second CABAC probability model being different from the first CABAC probability model;

encoding the first flag based on the first CABAC context and the first CABAC probability model when the block is in the AMVP mode and encoding the second flag based on the second CABAC context and the second CABAC probability model when the block is in the merge mode; and

encoding the block based on the first flag when the block is in the AMVP mode and encoding the block based the second flag when the block is in the merge mode.

8 . An apparatus, comprising:

one or more processors, wherein the one or more processors are configured for:

deriving a CABAC context index associated with a first CABAC context, the first CABAC context being associated with a first flag signaled at a block level of a video, the first flag indicating use of an affine motion model for the block when the block is in an AMVP mode;

the CABAC context index being associated with a second CABAC context, the second CABAC context being associated with a second flag signaled at the block level;

the second flag indicating use of an affine motion model for the block when the block is in a merge mode;

wherein the CABAC context index is derived using a same context derivation in the AMVP mode and in the merge mode;

the AMVP mode being a prediction mode wherein a motion vector difference with a motion predictor is coded for the block, and

the merge mode being a prediction mode wherein the motion vector difference is inferred to be zero for the block;

obtaining a first CABAC probability model associated with the first CABAC context, when the block is in the AMVP mode;

obtaining a second CABAC probability model associated with the second CABAC context, when the block is in the merge mode, the second CABAC probability model being different from the first CABAC probability model;

encoding the first flag based on the first CABAC context and the first CABAC probability model when the block is in the AMVP mode and encoding the second flag based on the second CABAC context and the second CABAC probability model when the block is in the merge mode; and

encoding the block based on the first flag when the block is in the AMVP mode and encoding the block based the second flag when the block is in the merge mode.

9 . The apparatus of claim 8 , wherein the merge mode comprises one of SbTMVP, mmvd, or DMVR.

10 . The method of claim 1 , wherein the context derivation derives the CABAC context index among three different context indices.

11 . The method of claim 1 , wherein the context derivation derives the CABAC context index by considering whether a left neighboring block uses an affine motion model and an above neighboring block uses an affine motion model.

12 . The method of claim 1 , wherein deriving the CABAC context index is based only on an availability or not of spatial neighbors.

13 . The method of claim 1 , further comprising constructing a virtual affine candidate to be considered when deriving the CABAC context index.

14 . The method of claim 13 , wherein constructing the virtual affine candidate is based on neighbor blocks coded in the AMVP mode and not using an affine motion model.

15 . The method of claim 14 , wherein the CABAC context index comprises one of:

0 if no inter neighbors are available;

1 if inter neighbors are available but no affine neighbors; or

2 if affine neighbors are available.

16 . The apparatus of claim 4 , wherein the context derivation derives the CABAC context index among three different context indices.

17 . The apparatus of claim 4 , wherein the context derivation derives the CABAC context index by considering whether a left neighboring block uses an affine motion model and an above neighboring block uses an affine motion model.

18 . The method of claim 7 , wherein the context derivation derives the CABAC context index among three different context indices.

19 . The method of claim 7 , wherein the context derivation derives the CABAC context index by considering whether a left neighboring block uses an affine motion model and an above neighboring block uses an affine motion model.

20 . The apparatus of claim 8 , wherein the context derivation derives the CABAC context index among three different context indices.

21 . The apparatus of claim 8 , wherein the context derivation derives the CABAC context index by considering whether a left neighboring block uses an affine motion model and an above neighboring block uses an affine motion model.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Apr 19, 2025
From: GALPIN, FRANCK; LE LEANNEC, FABRICE; BORDES, PHILIPPE
To: INTERDIGITAL VC HOLDINGS, INC.
Reel/Frame 070888/0850 →
Priority Claims (1)
EP 18306339 · Oct 10, 2018 · regional
Continuity (2)
Continuation 17282919
Related Publication 20240267522A1 · Aug 8, 2024
References Cited (42)
US 11323736B2 · Chen et al. · 2022 [cited by applicant]
US 11706417B2 · Chen · 2023 [cited by examiner]
US 20130004092A1 · Sasai et al. · 2013 [cited by applicant]
US 20130195199A1 · Guo et al. · 2013 [cited by applicant]
US 20170332095A1 · Zou et al. · 2017 [cited by applicant]
US 20180084260A1 · Chien et al. · 2018 [cited by applicant]
US 20180091816A1 · Chien et al. · 2018 [cited by applicant]
US 20180098063A1 · Chen et al. · 2018 [cited by applicant]
US 20190028731A1 · Chuang et al. · 2019 [cited by applicant]
US 20190058896A1 · Huang et al. · 2019 [cited by applicant]
US 20190104319A1 · Zhang et al. · 2019 [cited by applicant]
US 20190110064A1 · Zhang et al. · 2019 [cited by applicant]
US 20190335170A1 · Lee et al. · 2019 [cited by applicant]
US 20190335191A1 · Kondo · 2019 [cited by applicant]
US 20190342547A1 · Lee et al. · 2019 [cited by applicant]
US 20200260111A1 · Liu et al. · 2020 [cited by applicant]
US 20210185328A1 · Xu · 2021 [cited by examiner]
US 20210281838A1 · Lee et al. · 2021 [cited by applicant]
US 20220030269A1 · Laroche et al. · 2022 [cited by applicant]
US 20240357151A1 · Laroche · 2024 [cited by examiner]
AU 2018283967A1 · 2019 [cited by applicant]
CN 106559669A · 2017 [cited by applicant]
CN 108432250A · 2018 [cited by applicant]
CN 108605137A · 2018 [cited by applicant]
CN 112040247B · 2021 [cited by applicant]
GB 2579763A · 2020 [cited by examiner]
WO 2013001770A1 · 2013 [cited by applicant]
WO 2017054630A1 · 2017 [cited by applicant]
WO 2018128379A1 · 2018 [cited by applicant]
WO 2018128380A1 · 2018 [cited by applicant]
WO 2018131523A1 · 2018 [cited by applicant]
WO 2018174618A1 · 2018 [cited by applicant]
WO 2019070683A1 · 2019 [cited by applicant]
WO 2020052534A1 · 2020 [cited by applicant]
“Patdoc English Language Translation, WO 2018174618 A1”. [cited by applicant]
Bross, et al., “Versatile Video Coding (Draft 2); Document: JVET-K1001-v6”, Joint Video Experts Team (JVET) of ITU-T SG 16 WP 3 and ISO/IEC JTC 1/SC 29/WG11, 11th Meeting: Ljubljana, SI, Jul. 10-18, 2018, 140 pages. [cited by applicant]
Chen, et al., “Context Reduction for Inter and Split Syntax Elements”, Technicolor, JVET-N0600, Joint Video Experts Team (JVET) of ITU-T SG 16 WP 3 and ISO/IEC JTC 1/SC 29/WG 11, 14th Meeting: Geneva, CH, Mar. 19-27, 20… [cited by applicant]
Galpin (Technicolor), et al., “CE4-Related: Simplified Constructed Temporal Affine Merge Candidates”, JVET-L0522-v2, Macao, Oct. 5, 2018, 4 pages. [cited by applicant]
Interdigital, Inc., “Non-CE4: Affine and sub-block modes coding clean-up”, JVET of ITU-T SG16 WP3 and ISO/IEC JTC 1/SC29/WG11, 15th Meeting: Gothenburg, SE, Document: JVET-00500, Jul. 3-12, 2019, 4 pages. [cited by applicant]
ITU-T, “High Efficiency Video Coding”, H.265, Telecommunications Standardization Sector of ITU, Series H: Audiovisual and Multimedia Systems, Infrastructure of Audiovisual Services—Coding of Moving Video, Apr. 2015, 634… [cited by applicant]
Yang, et al., “CE4: Summary Report on Inter Prediction and Motion Vector Coding”, JVET-L0024-v2, Macao, China, Oct. 3-12, 2018, 48 pages. [cited by applicant]
Yang, et al., “Description of Corn Experiment 4 (CE4): Inter Prediction and Motion Vector Coding”, JVET-K1024-V3, Joint Video Experts Team (JVET) of ITU-T SG 16 WP 3 and ISO/IEC JTC 1 /SC 29/WG 11, 11th Meeting: Ljublja… [cited by applicant]