Deep Multimodal Clustering for Unsupervised Audiovisual Learning

TL;DR

This paper introduces DMC, an unsupervised multimodal clustering model that learns to associate audio and visual components, improving audiovisual understanding and sound localization without labeled data.

Contribution

The paper presents a novel unsupervised audiovisual learning model, DMC, that performs simultaneous clustering of audio and visual features for better correspondence learning.

Findings

DMC learns effective unimodal representations outperforming some classifiers.

DMC achieves state-of-the-art results in sound localization and multisource detection.

The model demonstrates strong performance in audiovisual understanding tasks.

Abstract

The seen birds twitter, the running cars accompany with noise, etc. These naturally audiovisual correspondences provide the possibilities to explore and understand the outside world. However, the mixed multiple objects and sounds make it intractable to perform efficient matching in the unconstrained environment. To settle this problem, we propose to adequately excavate audio and visual components and perform elaborate correspondence learning among them. Concretely, a novel unsupervised audiovisual learning model is proposed, named as \Deep Multimodal Clustering (DMC), that synchronously performs sets of clustering with multimodal vectors of convolutional maps in different shared spaces for capturing multiple audiovisual correspondences. And such integrated multimodal clustering network can be effectively trained with max-margin loss in the end-to-end fashion. Amounts of experiments in feature evaluation and audiovisual tasks are performed. The results demonstrate that DMC can learn effective unimodal representation, with which the classifier can even outperform human performance. Further, DMC shows noticeable performance in sound localization, multisource detection, and audiovisual understanding.