Learning to Synthesize Volumetric Meshes from Vision-based Tactile Imprints
This enables more precise contact information for robotic grasping and manipulation, though it is incremental as it builds on existing vision-based tactile sensing with new adaptation techniques.
The paper tackles the problem of reconstructing volumetric meshes of deformed elastomers from vision-based tactile sensor images, achieving accurate reconstruction across various domains as demonstrated by quantitative and qualitative results.
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.