Co-speech gesture generation on TED Expressive
0FGDGround Truth
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Ground Truth2023.03 | 0 | 0.703 | 178.827 | |
| DiffGesture2023.03 | 2.6 | 0.718 | 182.757 | |
| HA2G2023.03 | 5.306 | 0.641 | 173.899 | |
| Trimodal2023.03 | 12.613 | 0.563 | 154.088 | |
| Speech2Gesture2023.03 | 54.65 | 0.679 | 142.489 | |
| Attention Seq2Seq2023.03 | 54.92 | 0.152 | 122.693 | |
| Joint Embedding2023.03 | 64.555 | 0.13 | 120.627 |