Viewpoint Selection for Grasp Detection

TL;DR

This paper investigates how selecting optimal viewpoints can significantly enhance grasp detection accuracy and success rates in robotic manipulation, demonstrating that strategic viewpoint selection improves performance over random or head-on views.

Contribution

The paper introduces a viewpoint selection algorithm that improves grasp detection performance and success rates compared to baseline strategies.

Findings

Optimal viewpoints can dramatically increase detected grasps and accuracy.

Viewpoint selection improves grasp success rates by up to 12%.

Combining strategies yields the highest success improvements.

Abstract

In grasp detection, the robot estimates the position and orientation of potential grasp configurations directly from sensor data. This paper explores the relationship between viewpoint and grasp detection performance. Specifically, we consider the scenario where the approximate position and orientation of a desired grasp is known in advance and we want to select a viewpoint that will enable a grasp detection algorithm to localize it more precisely and with higher confidence. Our main findings are that the right viewpoint can dramatically increase the number of detected grasps and the classification accuracy of the top-n detections. We use this insight to create a viewpoint selection algorithm and compare it against a random viewpoint selection strategy and a strategy that views the desired grasp head-on. We find that the head-on strategy and our proposed viewpoint selection strategy can improve grasp success rates on a real robot by 8% and 4%, respectively. Moreover, we find that the combination of the two methods can improve grasp success rates by as much as 12%.