Social Interaction Video Generation on Social Interaction Benchmark
76.1Action AccuracySocialDirector
Evaluation results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| SocialDirectorBox Input=true, Backbone=Wan2.22026.05 | 76.1 | 72.3 | 88.2 | 4.29 | 16.2 | 16.8 | 43.7 | 85.1 | |
| GT OracleBox Input=false2026.05 | 75.9 | 72.6 | 90.8 | — | — | 15.6 | 39 | 88.8 | |
| SocialDirectorBox Input=true, Backbone=LTX-2.32026.05 | 74.3 | 70.8 | 88.3 | 5.89 | 17.7 | 16.4 | 32.4 | 86.7 | |
| Wan2.2Box Input=false2026.05 | 72.2 | 69.2 | 85.1 | 4.21 | 17.2 | 16.1 | 43.3 | 84.1 | |
| LTX-2.3Box Input=false2026.05 | 71.6 | 69.4 | 87.6 | 5.77 | 17.7 | 15.9 | 31.8 | 87 | |
| HunyuanVideoAvatarBox Input=false2026.05 | 68.9 | 68.6 | 88.5 | 5.06 | 18.8 | 16.5 | 36.2 | 86.8 | |
| MultiTalkBox Input=true2026.05 | 68.1 | 68.6 | 94.1 | 11.2 | 22.3 | 15 | 41.7 | 89.9 | |
| AnyTalkerBox Input=false2026.05 | 64.1 | 66.3 | 92.9 | 5.4 | 16.8 | 15.2 | 36.3 | 91.7 | |
| Bind-Your-AvatarBox Input=true2026.05 | 63.6 | 69.5 | 72.2 | 20.1 | 31.2 | 16.4 | 28.4 | 66.9 | |
| EchoMimic v3Box Input=false2026.05 | 60.1 | 65.5 | 98 | 5.26 | 17.6 | 14.4 | 35.9 | 90.8 |