DexArt: Benchmarking Generalizable Dexterous Manipulation with Articulated Objects
This work addresses the challenge of generalizable dexterous manipulation for robotics, but it is incremental as it builds on existing RL and representation learning methods.
The authors tackled the problem of enabling robots to manipulate diverse articulated objects using multi-finger hands, proposing the DexArt benchmark to evaluate policy generalizability on unseen objects, and achieved insights into how 3D representation learning affects RL decision-making with point cloud inputs.
To enable general-purpose robots, we will require the robot to operate daily articulated objects as humans do. Current robot manipulation has heavily relied on using a parallel gripper, which restricts the robot to a limited set of objects. On the other hand, operating with a multi-finger robot hand will allow better approximation to human behavior and enable the robot to operate on diverse articulated objects. To this end, we propose a new benchmark called DexArt, which involves Dexterous manipulation with Articulated objects in a physical simulator. In our benchmark, we define multiple complex manipulation tasks, and the robot hand will need to manipulate diverse articulated objects within each task. Our main focus is to evaluate the generalizability of the learned policy on unseen articulated objects. This is very challenging given the high degrees of freedom of both hands and objects. We use Reinforcement Learning with 3D representation learning to achieve generalization. Through extensive studies, we provide new insights into how 3D representation learning affects decision making in RL with 3D point cloud inputs. More details can be found at https://www.chenbao.tech/dexart/.