Adaptive Distillation Stabilizes One-Step Generative Video Compression
A compression-trained pixel diffusion codec gates misaligned DMD updates, cutting bitrate by 62.5% at matched LPIPS versus GLVC.
Jul 28, 20265 min2607.22772
2 articles on SOTA Papers
A compression-trained pixel diffusion codec gates misaligned DMD updates, cutting bitrate by 62.5% at matched LPIPS versus GLVC.
A two-stage VLA recipe combines 100K hours of UMI trajectories with cross-embodiment post-training, reaching 57.4% on RoboCasa365.
