Diffusion-based Inverse Observation Model for Artificial Skin

TL;DR

This paper introduces a diffusion-based inverse observation model that leverages tactile data from artificial skin to efficiently estimate object pose, addressing challenges of multimodal contact observations.

Contribution

It presents a novel diffusion model approach for tactile-based object pose estimation, improving sampling efficiency for multimodal contact hypotheses.

Findings

Efficient sampling of contact hypotheses demonstrated in simulated experiments.

The model effectively handles multimodal distributions in tactile sensing.

Enhanced accuracy in object pose estimation through diffusion-based inference.

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.