Reproduce “Sim-to-Real for Robotic Tactile Sensing via Physics-Based Simulation and Learned Latent Projections” on OpenResearch

Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.

Universal app for Apple silicon and Intel. Requires macOS 11 or later.

About this paper

Yashraj Narang, Balakumar Sundaralingam, Miles Macklin, Arsalan Mousavian, Dieter Fox

The authors build a tactile simulator that predicts BioTac electrical signals from simulated deformation. Isaac Gym models deformation, while variational autoencoders compress simulated meshes and electrode readings; learned projections connect their representations using paired examples. On a held-out ring indenter, the method reaches median RMS error 0.015, versus 0.016 for a supervised baseline, on one BioTac and this test split.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.