Text-to-motion generation on SnapMoGen (test) benchmark leaderboard · SOTA2 Research

Text-to-motion generation on SnapMoGen (test)

0.94R-Precision Top 1GT

Evaluation Results

Method	Links
GT2026.02		0.94	0.976	0.985	0.001	0.837	19.756	—
Real motions2026.05		0.94	0.976	0.985	0.001	0.837	—	—
TriC-Motion2026.02		0.907	0.969	0.985	26.346	0.675	19.831	—
ScaleMoGen2026.05		0.807	0.908	0.941	16.35	0.693	—	9.399
MoMask++2026.02		0.802	0.905	0.938	15.06	0.685	19.97	—
MoMask++2026.05		0.802	0.905	0.938	15.06	0.685	—	7.259
MoMask2026.02		0.777	0.888	0.927	17.404	0.664	—	—
MoMask2026.05		0.777	0.888	0.927	17.404	0.664	—	8.183
SALAD2026.05		0.742	0.875	0.922	23.248	0.658	—	9.238
StableMoFusion2026.02		0.679	0.823	0.888	27.801	0.605	—	—
StableMoFusion2026.05		0.679	0.823	0.888	27.801	0.605	—	9.064
MARDM2026.02		0.659	0.812	0.86	26.878	0.602	—	—
MARDM2026.05		0.659	0.812	0.86	26.878	0.602	—	9.812
MoGenTS2026.05		0.626	0.771	0.845	27.802	0.586	—	7.142
T2M-GPT2026.02		0.618	0.773	0.812	32.629	0.573	—	—
T2M-GPT2026.05		0.618	0.773	0.812	32.629	0.573	—	9.172
MDM2026.02		0.503	0.653	0.727	57.783	0.481	—	—
MDM2026.05		0.503	0.653	0.727	57.783	0.481	—	13.412