GUESS:GradUally Enriching SyntheSis for Text-Driven Human Motion Generation

Benchmark Model Rank Results
motion-synthesis-on-kit-motion-languageGUESS#14FID: 0.371R Precision Top3: 0.751Diversity: 10.933…
motion-synthesis-on-humanml3dGUESS#17FID: 0.109R Precision Top3: 0.787Diversity: 9.826…