Weight Averaging: A Simple Yet Effective Method to Overcome Catastrophic Forgetting in Automatic Speech Recognition

TL;DR

This paper introduces a straightforward weight averaging technique to mitigate catastrophic forgetting in automatic speech recognition models, enhancing continual learning across multiple tasks and languages.

Contribution

The paper proposes a simple weight averaging method, combined with knowledge distillation, to effectively prevent forgetting in end-to-end ASR models during adaptation.

Findings

Weight averaging improves performance on old and new tasks.

Method outperforms existing baselines in monolingual and multilingual ASR.

Simple approach achieves significant gains with minimal complexity.

Abstract

Adapting a trained Automatic Speech Recognition (ASR) model to new tasks results in catastrophic forgetting of old tasks, limiting the model's ability to learn continually and to be extended to new speakers, dialects, languages, etc. Focusing on End-to-End ASR, in this paper, we propose a simple yet effective method to overcome catastrophic forgetting: weight averaging. By simply taking the average of the previous and the adapted model, our method achieves high performance on both the old and new tasks. It can be further improved by introducing a knowledge distillation loss during the adaptation. We illustrate the effectiveness of our method on both monolingual and multilingual ASR. In both cases, our method strongly outperforms all baselines, even in its simplest form.