| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| User Study | PC-AVS | Lip Sync Quality4.46 | 18 | 1mo ago | |
| Conver-3D YouTube (test) | ViBES | FDD2.76 | 9 | 1mo ago | |
| Curated 5-Identity Audio-Visual Dataset (Macron, Paul, Obama, May, Stabenow) (test) | EmbedTalk | PSNR35.186 | 8 | 1mo ago | |
| NeRSemble v2 | VGM | PSNR21.586 | 6 | 1mo ago | |
| Self-Collected Dataset 50 identities | X-Portrait | FID32.77 | 6 | 1mo ago | |
| VFHQ (first 100 frames) | X-Portrait | FID26.22 | 6 | 1mo ago | |
| HDTF (test) | EchoMimic | IQA3.994 | 6 | 1mo ago | |
| CelebV-HQ (test) | Ours-Balanced | IQA3.588 | 6 | 1mo ago | |
| HDTF Set B (test) | SyncNet (Offset)-3 | 6 | 1mo ago | ||
| HDTF Set A (test) | SyncNet Offset-4 | 6 | 1mo ago | ||
| Talking Head Synthesis Datasets (test) | Narrating For You | PSNR35.94 | 5 | 1mo ago | |
| Talkinghead1kh | Face-V2V | PSNR24.17 | 5 | 1mo ago | |
| HDTF | Face-V2V | PSNR27 | 5 | 1mo ago |
