Learning to Synthesize Volumetric Meshes from Vision-based Tactile Imprints

TL;DR

This paper presents a method using graph neural networks to synthesize accurate volumetric meshes of elastomer deformation from vision-based tactile sensor images, enabling improved contact understanding for robotic manipulation.

Contribution

It introduces a supervised learning approach with self-supervised adaptation and augmentation techniques to transfer from simulation to real-world tactile sensing.

Findings

Accurately reconstructs elastomer deformation in real-world tactile sensors.

Effective transfer from simulation to real-world and across different sensors.

Improves contact modeling for robotic grasping and manipulation.

Abstract

Vision-based tactile sensors typically utilize a deformable elastomer and a camera mounted above to provide high-resolution image observations of contacts. Obtaining accurate volumetric meshes for the deformed elastomer can provide direct contact information and benefit robotic grasping and manipulation. This paper focuses on learning to synthesize the volumetric mesh of the elastomer based on the image imprints acquired from vision-based tactile sensors. Synthetic image-mesh pairs and real-world images are gathered from 3D finite element methods (FEM) and physical sensors, respectively. A graph neural network (GNN) is introduced to learn the image-to-mesh mappings with supervised learning. A self-supervised adaptation method and image augmentation techniques are proposed to transfer networks from simulation to reality, from primitive contacts to unseen contacts, and from one sensor to another. Using these learned and adapted networks, our proposed method can accurately reconstruct the deformation of the real-world tactile sensor elastomer in various domains, as indicated by the quantitative and qualitative results.