JVET-AF0058 AHG12: GOP-based RPR encoder control for ECM [K. Andersson, J. Ström, R. Yu, W. Ahmad, P. Wennersten (Ericsson)]

GOP-based RPR encoder control has been available for VTM since VTM-19.0. It selects coding in a reduced resolution when impact of re-scaling performance between source resolution and reduced resolution is expected to be limited. ECM does not have such functionality. This contribution proposes to add the GOP-based RPR encoder control from VTM to ECM to enable further experimentation of RPR functionality in ECM.

Objective performance (BDR) for RA 4K resolutions and QP 40, 43, 47 and 50 is reported as:

ECM-10.0 with GOP-based RPR versus ECM-10.0: Y: -3.25%, U 5.95%, V 5.15%

VTM-22.0 with GOP-based RPR versus VTM-22.0: Y: -8.92%, U -3.38%, V -5.87%

Objective performance (BDR) for RA all resolutions and QP 40, 43, 47 and 50 is reported as:

ECM-10.0 with GOP-based RPR versus ECM-10.0: Y: -1.48%, U 2.37%, V 1.54%

VTM-22.0 with GOP-based RPR versus VTM-22.0: Y: -4.08%, U -2.82%, V -3.92%

It is noted that the benefit of using the GOP-based RPR encoder control in ECM is less compared to its usage in VTM but that it still shows promising coding gains at challenging bitrates.

It is also asserted that subjective quality can be improved by the proposal.

The QP values experimented here are not in RPR CTC.

GOP-based RPR in VTM is provided as an optional encoder control but not enabled in RPR CTC in the context of VTM.

If performance gain is the goal, then CTC should be used to conduct experiments.

Further study was recommended.

CTC for EE2/ECM and general ECM improvements (7)

Contributions in this area were discussed at 1400–1610 on Tuesday 17 Oct. 2023 (chaired by Y. Ye).