Sample Efficient Grasp Learning Using Equivariant Models

TL;DR

This paper introduces an equivariant neural network approach for planar grasp detection that significantly improves sample efficiency, enabling effective grasp learning on a physical robot within a short time frame.

Contribution

It demonstrates that modeling the grasp function as SE(2)-equivariant enhances sample efficiency, allowing rapid learning of grasping skills with minimal data.

Findings

Achieved good grasp function approximation after only 600 attempts.

Enabled complete grasp learning on a physical robot in about 1.5 hours.

Showed that equivariant models improve data efficiency in robotic grasping.

Abstract

In planar grasp detection, the goal is to learn a function from an image of a scene onto a set of feasible grasp poses in $\mathrm{SE}(2)$. In this paper, we recognize that the optimal grasp function is $\mathrm{SE}(2)$-equivariant and can be modeled using an equivariant convolutional neural network. As a result, we are able to significantly improve the sample efficiency of grasp learning, obtaining a good approximation of the grasp function after only 600 grasp attempts. This is few enough that we can learn to grasp completely on a physical robot in about 1.5 hours.