Exploiting Parallel Audio Recordings to Enforce Device Invariance in CNN-based Acoustic Scene Classification

TL;DR

This paper introduces a novel domain adaptation method for acoustic scene classification that leverages parallel recordings from different devices to learn device-invariant features without requiring labels, improving robustness across device mismatches.

Contribution

It proposes an end-to-end domain-invariant classifier training approach using parallel audio recordings, eliminating the need for labeled data for domain adaptation.

Findings

Effective in learning device-invariant features

Reduces distribution mismatch impact

No labeled data required for adaptation

Abstract

Distribution mismatches between the data seen at training and at application time remain a major challenge in all application areas of machine learning. We study this problem in the context of machine listening (Task 1b of the DCASE 2019 Challenge). We propose a novel approach to learn domain-invariant classifiers in an end-to-end fashion by enforcing equal hidden layer representations for domain-parallel samples, i.e. time-aligned recordings from different recording devices. No classification labels are needed for our domain adaptation (DA) method, which makes the data collection process cheaper.