Classification based Grasp Detection using Spatial Transformer Network

TL;DR

This paper introduces a novel classification-based robotic grasp detection method utilizing multi-stage spatial transformer networks, achieving state-of-the-art accuracy and real-time performance while providing intermediate observability of grasp configurations.

Contribution

The paper presents a new classification-based approach with multi-stage STN that outperforms existing methods in accuracy and offers intermediate results for grasp configurations.

Findings

Achieved state-of-the-art accuracy in grasp detection

Operates in real-time with efficient computation

Provides intermediate observations of grasp configurations

Abstract

Robotic grasp detection task is still challenging, particularly for novel objects. With the recent advance of deep learning, there have been several works on detecting robotic grasp using neural networks. Typically, regression based grasp detection methods have outperformed classification based detection methods in computation complexity with excellent accuracy. However, classification based robotic grasp detection still seems to have merits such as intermediate step observability and straightforward back propagation routine for end-to-end training. In this work, we propose a novel classification based robotic grasp detection method with multiple-stage spatial transformer networks (STN). Our proposed method was able to achieve state-of-the-art performance in accuracy with real- time computation. Additionally, unlike other regression based grasp detection methods, our proposed method allows partial observation for intermediate results such as grasp location and orientation for a number of grasp configuration candidates.