Introduces a 2.7M-label benchmark for grasp feasibility from point clouds and shows a point-cloud transformer reaching 0.996 AUROC on novel objects while running faster than sampling planners.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.RO 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Learning Motion Feasibility from Point Clouds in Cluttered Environments
Introduces a 2.7M-label benchmark for grasp feasibility from point clouds and shows a point-cloud transformer reaching 0.996 AUROC on novel objects while running faster than sampling planners.