JVET-AL0148 AHG9: Generative face video and generative face video enhancement SEI messages for HEVC and AVC [J. Chen, B. Chen, Y. Ye (Alibaba), S. Gehlot, G.-M. Su, P. Yin, S. McCarthy (Dolby), A. Trioux, F. Yang (Xidian Univ.), Y.-K. Wang (Bytedance), H.-B. Teo, J.-Y. Thong, K. Abe (Panasonic), Y. Xu, K. Yang, Y. Li (SJTU)]
The generative face video and generative face video enhancement SEI messages were adopted into the draft text of VSEI version 4 at the 36th JVET meeting, and the corresponding interface text for VVC was adopted at 37th JVET meeting. This contribution proposes to add the generative face video and generative face video enhancement SEI messages to the next version of HEVC and AVC.
Software is intended to be provided by proponents. This will enable using to generate the content of the SEI message from the GFVC software, encode the base picture and write it along with the SEI message into the bitstream, decode the base picture and extract the parameters from the SEI message, and synthesize the face using the GFVC software.
It is expected that merge requests for the JM and HM software would be submitted before the finalization of the standards.
Decision: Adopt JVET-AL0148 for JVET-AL1006 and JVET-AL1017.
JVET-AL0148 AHG9: Generative face video and generative face video enhancement SEI messages for HEVC and AVC
JVET-AL0148 AHG9: Generative face video and generative face video enhancement SEI messages for HEVC and AVC