An investigation of modularity for noise robustness in conformer-based ASR

TL;DR

This paper explores how modularity in conformer-based ASR models can improve noise robustness and adaptation to new acoustic environments, highlighting challenges in environment classification.

Contribution

It investigates the role of modularity and environment awareness in conformer-based ASR, revealing limitations in environment classification for noise robustness.

Findings

Environment awareness improves performance in known environments.

Classifiers struggle to distinguish different noisy environments.

Simpler noise vs. clean speech distinction is more effective.

Abstract

Whilst state of the art automatic speech recognition (ASR) can perform well, it still degrades when exposed to acoustic environments that differ from those used when training the model. Unfamiliar environments for a given model may well be known a-priori, but yield comparatively small amounts of adaptation data. In this experimental study, we investigate to what extent recent formalisations of modularity can aid adaptation of ASR to new acoustic environments. Using a conformer based model and fixed routing, we confirm that environment awareness can indeed lead to improved performance in known environments. However, at least on the (CHIME) datasets in the study, it is difficult for a classifier module to distinguish different noisy environments, a simpler distinction between noisy and clean speech being the optimal configuration. The results have clear implications for deploying large models in particular environments with or without a-priori knowledge of the environmental noise.