Multi-level Causal LLM-based Text-to-Motion Generation with Human Alignment

Open paper
Benchmark Model Rank Results
motion-synthesis-on-humanml3dMoTiGA (Llama)–FID: 0.041R Precision Top3: 0.808Diversity: 9.659
motion-synthesis-on-kit-motion-languageMoTiGA (Llama)–FID: 0.18R Precision Top3: 0.79Diversity: 10.956

Rank counts only results with a code link.