Keith-Luo/pick_and_place_occlusion

pick_and_place QuestArmTeleop LeRobot v3 dataset. RGB is encoded as LeRobot video and lossless aligned depth is retained under rgbd/. Point cloud Point clouds are stored directly in Parquet as observation.pointcloud.xyz (float32 metres, shape [1024, 3], base_link), observation.pointcloud.rgb (uint8 RGB), and observation.pointcloud.valid_mask. They share the same row, frame index and timestamp as state/action. See conversion_info.json for coordinate frames, units… See the full description on the dataset page: https://huggingface.co/datasets/Keith-Luo/pick_and_place_occlusion.

<iframe src="https://silytics.si/embed/item/bf2d0fa0-1567-4e5d-a49b-c1a912e6545d" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 18 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.