Back to Search Document details
13th Meeting: Marrakech, January 2019 2019-01-15 12:58
CE4-related: HMVP and parallel processing with tiles and tile groups
Authors: Anand Meher Kotra Huawei Jie Chen Huawei Biao Wang Huawei S. Esenlik Huawei Han Gao Huawei
Abstract
In VTM-3.0, History-based motion vector prediction (HMVP) FIFO table is reset for each CTU Row to enable parallel processing for tools such as Wavefronts (WPP). From VTM-3.1 onwards the concept of slices is replaced with tile groups. Therefore this contribution extends the technique of re-setting the HMVP table for Tiles and Tile groups. Specifically, the HMVP table is re-set at the beginning of a new Tile group or a new Tile and also at the beginning of each CTU row inside a Tile.
Meeting Notes
JVET-M0300 CE4-related: HMVP and parallel processing with tiles and tile groups [A. M. Kotra, J. Chen, B. Wang, S. Esenlik, H. Gao (Huawei)]

Notes from BoG report review: This resets the history table at the beginning of each tile, however under the assumption that tile boundaries coincide with CTU boundaries which may not be the case with flexible tile concepts. It was suggested that proponents should communicate with tile experts and see whether there is a misaligment with concepts that will be put into VVC draft 4). It was later confirmed by the text editor that only a small (kind of editorial) change would be required. (might be that depending on further discussion on tile concepts; some more alignments may be necessary).

JVET-M0300 CE4-related: HMVP and parallel processing with tiles and tile groups [A. M. Kotra, J. Chen, B. Wang, S. Esenlik, H. Gao (Huawei)]

Notes from BoG report review: This resets the history table at the beginning of each tile, however under the assumption that tile boundaries coincide with CTU boundaries which may not be the case with flexible tile concepts. It was suggested that proponents should communicate with tile experts and see whether there is a misaligment with concepts that will be put into VVC draft 4). It was later confirmed by the text editor that only a small (kind of editorial) change would be required. (might be that depending on further discussion on tile concepts; some more alignments may be necessary).

References:
JVET-L0106 JVET-L0158 JVET-L0226 JVET-L0686 JVET-M300
Cited:
JVET-M0012 JVET-M0024 JVET-M1001 JVET-N0002 JVET-N0075 JVET-N0076 JVET-N0077 JVET-N0078 JVET-N0079 JVET-N0081 JVET-N0085 JVET-N0086 JVET-N0088 JVET-N0097 JVET-N0107 JVET-N0108 JVET-N0115 JVET-N0116 JVET-N0119 JVET-N0121 more than 20
PATENTS:
AU2020276527B9 1.00 2024-06-27 US11985305B2 1.00 2024-05-14 US12401784B2 1.00 2025-08-26 US20230007273A1 1.00 2023-01-05 JP2023038228A 0.76 2023-03-16 ES2946909T3 0.34 2023-07-27 US12015797B2 0.30 2024-06-18 US20210400304A1 0.24 2021-12-23
ARCHIVE v3 :
JVET-M0300-v3-clean.docx JVET-M0300-v3.docx JVET-M0300-v3.pptx
Decisions
JVET-M0300 CE4-related: HMVP and parallel processing with tiles and tile groups
pending
Notes from BoG report review: This resets the history table at the beginning of each tile, however under the assumption that tile boundaries coincide with CTU boundaries which may not be the case with flexible tile concepts. It was suggested that proponents should communicate with tile experts and see whether there is a misaligment with concepts that will be put into VVC draft 4). It was later confirmed by the text editor that only a small (kind of editorial) change would be required. (might be that depending on further discussion on tile concepts; some more alignments may be necessary).
Citation