Dex-Net 2.0

Dex-Net 2.0 is a grasping robotics dataset maintained by UC Berkeley AUTOLAB. Primary hardware platform: synthetic point clouds, parallel-jaw grasps. Access is open under not stated.

Dataset specifications
Year2017
Embodimentssynthetic point clouds, parallel-jaw grasps
Task categoriespick-and-place
Licensenot stated
Accessopen
MaintainerUC Berkeley AUTOLAB