GraspXL: Generating Grasping Motions for Diverse Objects at Scale

arXiv:2403.19649v1 Announce Type: cross
Abstract: Human hands possess the dexterity to interact with diverse objects such as grasping specific parts of the objects and/or approaching them from desired directions. More importantly, humans can grasp objects of any shape without object-specific skills. Recent works synthesize grasping motions following single objectives such as a desired approach heading direction or a grasping area. Moreover, they usually rely on expensive 3D hand-object data during training and inference, which limits their capability to synthesize grasping …

abstract arxiv cs.cv cs.ro diverse grasping human humans object objects scale skills them type