The Feeling of Success: Does Touch Sensing Help Predict Grasp Outcomes?

TL;DR

This study investigates whether touch sensing, combined with vision, improves the prediction of grasp success in robotic manipulation, demonstrating that tactile data significantly enhances outcome prediction accuracy.

Contribution

The paper presents a multimodal deep learning approach using GelSight tactile sensors and vision to predict grasp outcomes, showing tactile data's added value.

Findings

Tactile sensing improves grasp outcome prediction accuracy.

Combining vision and touch yields better predictions than using either modality alone.

Over 9,000 grasp trials validate the effectiveness of the multimodal approach.

Abstract

A successful grasp requires careful balancing of the contact forces. Deducing whether a particular grasp will be successful from indirect measurements, such as vision, is therefore quite challenging, and direct sensing of contacts through touch sensing provides an appealing avenue toward more successful and consistent robotic grasping. However, in order to fully evaluate the value of touch sensing for grasp outcome prediction, we must understand how touch sensing can influence outcome prediction accuracy when combined with other modalities. Doing so using conventional model-based techniques is exceptionally difficult. In this work, we investigate the question of whether touch sensing aids in predicting grasp outcomes within a multimodal sensing framework that combines vision and touch. To that end, we collected more than 9,000 grasping trials using a two-finger gripper equipped with GelSight high-resolution tactile sensors on each finger, and evaluated visuo-tactile deep neural network models to directly predict grasp outcomes from either modality individually, and from both modalities together. Our experimental results indicate that incorporating tactile readings substantially improve grasping performance.