DexArt: Benchmarking Generalizable Dexterous Manipulation with Articulated Objects

(a) DexArt Task Suite (b) Seen Objects (c) Unseen Objects Figure 1. Overview. (a) We propose DexArt, a task suite of Dexterous manipulation with Articulated object using point cloud observation. (b) We experiment with extensive benchmark methods that learn category-level manipulation policy on seen objects. (c) We evaluate the policies' generalizability on a collection of unseen objects, as well as their robustness to camera viewpoint change.