English

Diffusion-based Inverse Observation Model for Artificial Skin

Robotics 2025-06-18 v1

Abstract

Contact-based estimation of object pose is challenging due to discontinuities and ambiguous observations that can correspond to multiple possible system states. This multimodality makes it difficult to efficiently sample valid hypotheses while respecting contact constraints. Diffusion models can learn to generate samples from such multimodal probability distributions through denoising algorithms. We leverage these probabilistic modeling capabilities to learn an inverse observation model conditioned on tactile measurements acquired from a distributed artificial skin. We present simulated experiments demonstrating efficient sampling of contact hypotheses for object pose estimation through touch.

Keywords

Cite

@article{arxiv.2506.13986,
  title  = {Diffusion-based Inverse Observation Model for Artificial Skin},
  author = {Ante Maric and Julius Jankowski and Giammarco Caroleo and Alessandro Albini and Perla Maiolino and Sylvain Calinon},
  journal= {arXiv preprint arXiv:2506.13986},
  year   = {2025}
}

Comments

Accepted to RSS 2025 workshop on Navigating Contact Dynamics in Robotics

R2 v1 2026-07-01T03:20:42.453Z