STReSSD: Sim-To-Real from Sound for Stochastic Dynamics

TL;DR

This paper introduces STReSSD, a framework that uses sound to bridge the simulation-to-reality gap for stochastic dynamics, demonstrated with a bouncing ball, enabling better prediction and control in robotic tasks.

Contribution

It presents a novel sound-based inference framework and stochastic simulator calibration method for dynamic physical interactions, specifically applied to bouncing ball scenarios.

Findings

Sound-based inference accurately estimates stochastic parameters.

Calibrated models improve prediction of ball dynamics.

Audio perception enables real-time robotic interaction with bouncing objects.

Abstract

Sound is an information-rich medium that captures dynamic physical events. This work presents STReSSD, a framework that uses sound to bridge the simulation-to-reality gap for stochastic dynamics, demonstrated for the canonical case of a bouncing ball. A physically-motivated noise model is presented to capture stochastic behavior of the balls upon collision with the environment. A likelihood-free Bayesian inference framework is used to infer the parameters of the noise model, as well as a material property called the coefficient of restitution, from audio observations. The same inference framework and the calibrated stochastic simulator are then used to learn a probabilistic model of ball dynamics. The predictive capabilities of the dynamics model are tested in two robotic experiments. First, open-loop predictions anticipate probabilistic success of bouncing a ball into a cup. The second experiment integrates audio perception with a robotic arm to track and deflect a bouncing ball in real-time. We envision that this work is a step towards integrating audio-based inference for dynamic robotic tasks. Experimental results can be viewed at https://youtu.be/b7pOrgZrArk.