Damage Control During Domain Adaptation for Transducer Based Automatic Speech Recognition
This addresses the issue of maintaining performance on original domains during adaptation for speech recognition systems, but it is incremental as it builds on existing adapter methods.
The paper tackles the problem of catastrophic forgetting during domain adaptation for automatic speech recognition, proposing techniques like limited training and regularized adapter modules to adapt models to new domains while limiting degradation on the original domain, achieving strong results on target domains with controlled degradation.
Automatic speech recognition models are often adapted to improve their accuracy in a new domain. A potential drawback of model adaptation to new domains is catastrophic forgetting, where the Word Error Rate on the original domain is significantly degraded. This paper addresses the situation when we want to simultaneously adapt automatic speech recognition models to a new domain and limit the degradation of accuracy on the original domain without access to the original training dataset. We propose several techniques such as a limited training strategy and regularized adapter modules for the Transducer encoder, prediction, and joiner network. We apply these methods to the Google Speech Commands and to the UK and Ireland English Dialect speech data set and obtain strong results on the new target domain while limiting the degradation on the original domain.