GraspLDP: Towards Generalizable Grasping Policy via Latent Diffusion

TL;DR

This paper introduces GraspLDP, a latent diffusion policy framework that incorporates grasp priors and self-supervised learning to improve the precision and generalization of robotic grasping in manipulation tasks.

Contribution

It proposes a novel latent diffusion policy with grasp priors and self-supervised reconstruction, enhancing grasping accuracy and generalization in robotic manipulation.

Findings

Outperforms baseline methods in simulation and real robot experiments.

Demonstrates strong dynamic grasping capabilities.

Achieves better generalization across objects and environments.

Abstract

This paper focuses on enhancing the grasping precision and generalization of manipulation policies learned via imitation learning. Diffusion-based policy learning methods have recently become the mainstream approach for robotic manipulation tasks. As grasping is a critical subtask in manipulation, the ability of imitation-learned policies to execute precise and generalizable grasps merits particular attention. Existing imitation learning techniques for grasping often suffer from imprecise grasp executions, limited spatial generalization, and poor object generalization. To address these challenges, we incorporate grasp prior knowledge into the diffusion policy framework. In particular, we employ a latent diffusion policy to guide action chunk decoding with grasp pose prior, ensuring that generated motion trajectories adhere closely to feasible grasp configurations. Furthermore, we introduce a self-supervised reconstruction objective during diffusion to embed the graspness prior: at each reverse diffusion step, we reconstruct wrist-camera images back-projected the graspness from the intermediate representations. Both simulation and real robot experiments demonstrate that our approach significantly outperforms baseline methods and exhibits strong dynamic grasping capabilities.