SoundSpaces 2.0: A Simulation Platform for Visual-Acoustic Learning

TL;DR

SoundSpaces 2.0 is a high-fidelity, real-time geometry-based audio simulation platform for 3D environments, enabling advanced audio-visual research and embodied learning tasks.

Contribution

It introduces a novel, fast, and realistic acoustic simulation platform that supports continuous spatial sampling and generalizes to new environments, advancing audio-visual research capabilities.

Findings

Benchmark against real-world audio measurements shows high fidelity.

Demonstrated improved performance in embodied navigation tasks.

Showcased effective sim2real transfer in automatic speech recognition.

Abstract

We introduce SoundSpaces 2.0, a platform for on-the-fly geometry-based audio rendering for 3D environments. Given a 3D mesh of a real-world environment, SoundSpaces can generate highly realistic acoustics for arbitrary sounds captured from arbitrary microphone locations. Together with existing 3D visual assets, it supports an array of audio-visual research tasks, such as audio-visual navigation, mapping, source localization and separation, and acoustic matching. Compared to existing resources, SoundSpaces 2.0 has the advantages of allowing continuous spatial sampling, generalization to novel environments, and configurable microphone and material properties. To our knowledge, this is the first geometry-based acoustic simulation that offers high fidelity and realism while also being fast enough to use for embodied learning. We showcase the simulator's properties and benchmark its performance against real-world audio measurements. In addition, we demonstrate two downstream tasks -- embodied navigation and far-field automatic speech recognition -- and highlight sim2real performance for the latter. SoundSpaces 2.0 is publicly available to facilitate wider research for perceptual systems that can both see and hear.