Weight Averaging: A Simple Yet Effective Method to Overcome Catastrophic Forgetting in Automatic Speech Recognition
This addresses the problem of continual learning in ASR for applications like adapting to new speakers or languages, though it is incremental as it builds on existing weight averaging techniques.
The paper tackles catastrophic forgetting in automatic speech recognition when adapting models to new tasks, proposing weight averaging to maintain performance on both old and new tasks, achieving strong outperformance over baselines in monolingual and multilingual settings.
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.