GRAB benchmark shows object quality as the dominant factor in grasp success across gripper types, with physical interaction constraints causing most failures in cluttered food waste sorting.
Dex-net 3.0: Computing robust vacuum suction grasp targets in point clouds using a new analytic model and deep learning
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.RO 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
A Real-World Grasping-in-Clutter Performance Evaluation Benchmark for Robotic Food Waste Sorting
GRAB benchmark shows object quality as the dominant factor in grasp success across gripper types, with physical interaction constraints causing most failures in cluttered food waste sorting.