Learning to Localize Sound Source in Visual Scenes

TL;DR

This paper introduces an unsupervised learning approach for localizing sound sources in visual scenes using a two-stream network with attention, demonstrating effectiveness and adaptability with minimal supervision.

Contribution

The paper presents a novel unsupervised algorithm for sound source localization in visual scenes, with a flexible architecture extendable to supervised and semi-supervised learning.

Findings

Unsupervised method can sometimes produce false conclusions.

Minimal supervision helps correct false conclusions.

Effective localization achieved with limited supervision.

Abstract

Visual events are usually accompanied by sounds in our daily lives. We pose the question: Can the machine learn the correspondence between visual scene and the sound, and localize the sound source only by observing sound and visual scene pairs like human? In this paper, we propose a novel unsupervised algorithm to address the problem of localizing the sound source in visual scenes. A two-stream network structure which handles each modality, with attention mechanism is developed for sound source localization. Moreover, although our network is formulated within the unsupervised learning framework, it can be extended to a unified architecture with a simple modification for the supervised and semi-supervised learning settings as well. Meanwhile, a new sound source dataset is developed for performance evaluation. Our empirical evaluation shows that the unsupervised method eventually go through false conclusion in some cases. We show that even with a few supervision, false conclusion is able to be corrected and the source of sound in a visual scene can be localized effectively.