Neural Grasp Distance Fields for Robot Manipulation

TL;DR

This paper introduces Neural Grasp Distance Fields (NGDF), a neural representation that predicts a continuous grasp cost for robot manipulation, enabling joint optimization of grasping and motion planning with improved success rates.

Contribution

The paper presents NGDF, a novel neural field that models grasp quality as a continuous distance function, allowing seamless integration into trajectory optimization for robotic grasping.

Findings

Outperforms baselines by 63% in success rate

Generalizes to unseen objects and poses

Enables joint optimization of grasp and motion planning

Abstract

We formulate grasp learning as a neural field and present Neural Grasp Distance Fields (NGDF). Here, the input is a 6D pose of a robot end effector and output is a distance to a continuous manifold of valid grasps for an object. In contrast to current approaches that predict a set of discrete candidate grasps, the distance-based NGDF representation is easily interpreted as a cost, and minimizing this cost produces a successful grasp pose. This grasp distance cost can be incorporated directly into a trajectory optimizer for joint optimization with other costs such as trajectory smoothness and collision avoidance. During optimization, as the various costs are balanced and minimized, the grasp target is allowed to smoothly vary, as the learned grasp field is continuous. We evaluate NGDF on joint grasp and motion planning in simulation and the real world, outperforming baselines by 63% execution success while generalizing to unseen query poses and unseen object shapes. Project page: https://sites.google.com/view/neural-grasp-distance-fields.