-
Add observation.objects.cube [32] + rename observation_vector->observation.state (non-destructive superset)
Per-frame cube object-state (R/G/B/P x 2 cams x [cx/W,cy/H,sqrt(area/TYP),soft_vis]) via per-episode color seg; hold-last+vis=0 on occlusion; absent color=zeros. observation_vector renamed to lerobot-canonical observation.state (values identical). Cameras/action/indices/videos unchanged. Stats + PROCESSING.md/README updated.
-
Upload folder using huggingface_hub (part 2)
