Action-conditioned Video Generation on RoboNet benchmark leaderboard · SOTA2 Research

Action-conditioned Video Generation on RoboNet

65.8FVDiVideoGPT

Evaluation Results

Method	Links
iVideoGPT2025.10		65.8	27.2	89.5	0.053	5.4
RoDyn2025.10		67.6	28	92	0.032	3.1
GHVAE2025.10		95.2	24.7	89.1	0.036	—
SVG2025.10		123.2	23.9	87.8	0.06	—
MaskViT2025.10		133.5	23.2	80.5	0.042	—