Speech-driven Holistic Expression and Gesture Generation on SHOW 2023 (test) benchmark leaderboard · SOTA2 Research

Speech-driven Holistic Expression and Gesture Generation on SHOW 2023 (test)

0.0018FMDDiffSHEG

Evaluation Results

Method	Links
DiffSHEG2024.01		0.0018	0.0016	0.923	0.0027	0.902	0.912	0.81
TalkShow2024.01		0.0022	0.0023	0.74	0.0032	0.869	0.902	0.703
LS3DCG2024.01		0.0023	0.0023	0.708	0.0048	0.947	0.981	0.645
Ours w/o E0(t)2024.01		0.0025	0.0025	0.728	0.0031	0.892	0.927	0.694
Ours Naïve Concat2024.01		0.0026	0.0019	0.766	0.0041	0.896	0.932	0.618
Ours w/o Detach2024.01		0.0026	0.0023	0.745	0.0038	0.899	0.926	0.673
Ours Reverse Direction2024.01		0.0027	0.0027	0.731	0.0035	0.882	0.926	0.695
TalkShow2024.01		0.0028	0.0041	0.618	0.0033	0.872	0.894	0.711