DistantSky/placement_visual_conditioned

Robotics Dataset

This dataset contains 197 episodes of a Hessian dual-arm robot performing a visual-conditioned placement task, 'place the holder', with 66,397 frames of RGB video from three cameras, state/action data, and task instructions.

Catalog metadata

Repository owner: DistantSky
Source: Hugging Face / LeRobot
Robot: hessian
Episodes: 197
Frames: 66397
Cameras: 3
Frames per second: 60
License: apache-2.0

Example tasks

place the holder

Browse the catalog to compare robots, tasks, sensors, and licenses. Sign in to open protected download and model links.