Back to Search
Document details
CE10.2: Generalized OBMC (Test 10.2.3)
Abstract
This contribution reports the results of generalized OBMC as described in JVET-J0022 [1], implemented in the JVET VTM-1.0 and the BMS-1.0. The proposed generalized OBMC gives, in the VTM code, an average luma BD-rate gain of -1.01% in RA configuration and -1.40% in LDB configuration, for 14% and 17% encoder, and 25% and 20% decoder complexity increase respectively. In the BMS code, it leads to an average luma BD-rate gain of -1.24% in RA and -1.94% in LDB, for 7% and 9% encoder, and 31% and 36% decoder complexity increase respectively.
JVET-K0213 CE10.2: Generalized OBMC (Test 10.2.3) [A. Robert, T. Poirier, F. Le Léannec (Technicolor)]
Non-adjacent spatial candidates are added.
- Derivation of new candidates
- The search grid is based on block width and block height, with maximum search range 96
- Total (96 / max (width, height)) search points are checked. The detailed search pattern for each round is described in the figure above.
- When max (width, height) is greater than the threshold (64), the search grid is 32x32
- Redundancy checking (in a different way for other merge candidates pruning) is performed for the added merge candidates.
- New candidates are added after TMVP candidates in the merge candidate list.
- Maximum merge candidate number is 10 or 8 in VTM 1.0 and 11 or 9 in BMS 1.0.
References:
Cited:
Decisions
merged
Maximum merge candidate number is 10 or 8 in VTM 1.0 and 11 or 9 in BMS 1.0.
Citation