DexDiffuser: Generating Dexterous Grasps with Diffusion Models
This addresses the challenge of robust multi-finger grasping for robotics, offering a significant but incremental improvement over existing methods.
The paper tackles the problem of generating dexterous grasps for objects from partial point clouds, introducing DexDiffuser, which achieves a 9.12% higher grasp success rate in simulation and 19.44% higher in real robot experiments compared to the state-of-the-art method FFHNet.
We introduce DexDiffuser, a novel dexterous grasping method that generates, evaluates, and refines grasps on partial object point clouds. DexDiffuser includes the conditional diffusion-based grasp sampler DexSampler and the dexterous grasp evaluator DexEvaluator. DexSampler generates high-quality grasps conditioned on object point clouds by iterative denoising of randomly sampled grasps. We also introduce two grasp refinement strategies: Evaluator-Guided Diffusion (EGD) and Evaluator-based Sampling Refinement (ESR). The experiment results demonstrate that DexDiffuser consistently outperforms the state-of-the-art multi-finger grasp generation method FFHNet with an, on average, 9.12% and 19.44% higher grasp success rate in simulation and real robot experiments, respectively. Supplementary materials are available at https://yulihn.github.io/DexDiffuser_page/