Back to Search
Document details
AHG14: Performance Evaluation by CTU based Intra Refresh
Abstract
This contribution reports the performance evaluation by CTU based intra refresh approach as information. While the anchor condition candidate in JVET-L0160 used a CU line as a granularity of the refresh area, this contribution used a CTU line. For 1080p content, CTU size is set to 64 instead of 128 to realize that the intra refresh period becomes 30 frames. The preliminary simulation results shows 7.4% Y-BD rate loss compared to with / without intra refresh coding condition. This contribution also suggests to change the intra refresh period to 30 frame, instead of 32 frame.
JVET-M0871 AHG14: Performance Evaluation by CTU based Intra Refresh [K. Kawamura, S. Naito (KDDI)] [late]
This contribution was presented Thursday 17 January at 1830 (chaired by F. Bossen).
This contribution reports a performance evaluation of a CTU-based intra refresh approach as information. While the anchor condition candidate in JVET-L0160 used a CU line as a granularity of the refresh area, this contribution used a CTU line. For 1080p content, CTU size was set to 64 instead of 128 to produce an intra refresh period of 30 frames. The preliminary simulation results reportedly showed 7.4% Y-BD rate loss compared to with/without the intra refresh coding condition. This contribution also suggests to change the intra refresh period to 30 frames instead of 32 frame.
This was presented for information only. No action was taken on it.
Adaptive streaming and cross-RAP referencing
References:
Decisions
Adaptive streaming and cross-RAP referencing
Citation