← Back to Benchmarks
simmediummanipulation-datametric · varies
RoboPCA: Pose-centered Affordance Learning from Human Demonstrations for Robot Manipulation
Description
Understanding spatial affordances -- comprising the contact regions of object interaction and the corresponding contact poses -- is essential for robots to effectively manipulate objects and accomplish diverse tasks. However, existing spatial affordance prediction methods mainly focus on locating the contact regions while delegating the pose to independent pose estimation approaches, which can lead to task failures due to inconsistencies between predicted contact regions and candidate poses. In