Plan, Don't Pose: Long Composite Motion Generation with Text-Aligned BFM
Open paper
Benchmark
Model
Rank
Results
motion-synthesis-on-humanml3d
Text2BFM
–
FID: 1.172
R Precision Top3: 0.876
Multimodality: 1.233
motion-synthesis-on-kit-motion-language
Text2BFM
–
FID: 1.482
R Precision Top3: 0.901
Multimodality: 1.367
Rank counts only results with a code link.