Conversational Behavior Generation on SynMSI (test)
3Semantic AlignmentViBES
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| ViBEScategory=Speech–Language–Behavior Unified Models2025.12 | 3 | 0.55 | 3.5 | |
| MoMaskcategory=Text-to-motion generation2025.12 | 2.6 | 0.4 | 2.9 | |
| T2Mcategory=Text-to-motion generation2025.12 | 2 | 0.3 | 2.4 | |
| MotionGPTcategory=Text-to-motion generation2025.12 | 1.9 | 0.3 | 2.1 | |
| LOMcategory=Co-speech gesture generation2025.12 | 1.7 | 0.2 | 2.3 | |
| EMAGEcategory=Co-speech gesture generation2025.12 | 1.6 | 0.1 | 2 | |
| SynTalkercategory=Co-speech gesture generation2025.12 | 1.5 | 0.2 | 2.2 |