Core Concepts
Proposing a dynamic grasping pipeline with a learned meta-controller to improve success rates and reduce grasping time.
Stats
"Our experiments show the meta-controller improves the grasping success rate (up to 28% in the most cluttered environment) and reduces grasping time, compared to the strongest baseline."
"Despite being trained only with 3-6 random cuboidal obstacles, our meta-controller generalizes well to 7-9 obstacles and more realistic out-of-domain household setups with unseen obstacle shapes."
Quotes
"Our meta-controller learns to reason about the reachable workspace and maintain the predicted pose within the reachable region."
"It learns to generate a small look-ahead time when the predicted trajectory is not accurate."
"It learns to produce a small but sufficient time budget for motion planning."