| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Music-to-Dance | AIST++ | FIDk17.1 | 17 | |
| Dance-to-Music | AIST++ | BCS98.6 | 17 | |
| Music-to-Dance Synthesis | AIST++ (test) | FID (k)17.1 | 16 | |
| Conditional Motion Generation | AIST++ (test) | Beat Alignment Score0.292 | 16 | |
| Human Motion Prediction | AIST++ 10 dance genres | Break Genre Score0.037 | 14 | |
| Music-to-dance generation | AIST++ (val and test) | FIDk15.57 | 14 | |
| Dance-to-Music | AIST++ (test) | BCS97.64 | 11 | |
| Dance Generation | AIST++ (test) | FID0.4988 | 7 | |
| Audio-to-video generation (A2V) | AIST++ (test) | FVD38.04 | 6 | |
| Human Pose Lifting | AIST++ | MPJPE104.8 | 6 | |
| Music-conditioned Dance Generation | AIST++ (test) | FIDg12.96 | 6 | |
| Music-to-Dance Generation | AIST++ (test) | PFC1.1722 | 5 | |
| Music-driven Dance Generation | AIST++ (test) | Distk5.94 | 5 | |
| 3D Human Motion Capture | AIST++ (test) | MPJPE33.3 | 5 | |
| Music-to-dance generation | AIST++ | Elo1,751 | 5 | |
| Vector Quantization / Latent Space Modeling | AIST++ standard (val) | Activation Percentage72 | 4 | |
| 2D Dance Pose Generation | AIST++2D proportion-aligned (test) | FID29.31 | 4 | |
| Video-to-music generation | AIST++ | BCS1 | 4 | |
| Body Mesh Recovery | AIST++ 27 | MPJPE64.3 | 4 | |
| Music Generation | AIST++ | OVL4.3 | 3 | |
| Video-to-audio generation (V2A) | AIST++ (test) | FAD1.11 | 2 | |
| Text-to-Dance | AIST++ | R-TOP10.588 | 2 | |
| Novel View Synthesis | AIST++ S21 sequence | LPIPS0.205 | 2 | |
| Novel View Synthesis | AIST++ (S13 sequence) | LPIPS0.183 | 2 | |
| Audio-Video reconstruction | AIST++ (test) | FAD0.9 | 1 |