# Audio-Visual Model Distillation Using Acoustic Images

## TL;DR

This paper introduces a multimodal approach to improve audio classification by distilling knowledge from visual and acoustic image data, resulting in more robust and generalizable audio representations.

## Contribution

It presents a novel method of using acoustic images and visual data for audio model distillation, enhancing robustness over traditional single-microphone models.

## Key findings

- Distilled models outperform single-microphone models in robustness.
- Multimodal training improves generalization in audio classification.
- Acoustic images provide valuable spatial information for audio understanding.

## Abstract

In this paper, we investigate how to learn rich and robust feature representations for audio classification from visual data and acoustic images, a novel audio data modality. Former models learn audio representations from raw signals or spectral data acquired by a single microphone, with remarkable results in classification and retrieval. However, such representations are not so robust towards variable environmental sound conditions. We tackle this drawback by exploiting a new multimodal labeled action recognition dataset acquired by a hybrid audio-visual sensor that provides RGB video, raw audio signals, and spatialized acoustic data, also known as acoustic images, where the visual and acoustic images are aligned in space and synchronized in time. Using this richer information, we train audio deep learning models in a teacher-student fashion. In particular, we distill knowledge into audio networks from both visual and acoustic image teachers. Our experiments suggest that the learned representations are more powerful and have better generalization capabilities than the features learned from models trained using just single-microphone audio data.