Fast Text-Only Domain Adaptation of RNN-Transducer Prediction Network

TL;DR

This paper demonstrates a fast, effective method for domain adaptation of RNN-Transducer speech recognition models using only small amounts of textual data, avoiding complex fusion methods and external language models.

Contribution

It introduces a novel approach to adapt RNN-Transducer models with minimal data by leveraging the prediction network as a language model, simplifying the adaptation process.

Findings

Achieves 10-45% relative WER reduction across multiple tasks.

Enables quick adaptation without external language models.

Provides insights into the prediction network's language modeling capabilities.

Abstract

Adaption of end-to-end speech recognition systems to new tasks is known to be challenging. A number of solutions have been proposed which apply external language models with various fusion methods, possibly with a combination of two-pass decoding. Also TTS systems have been used to generate adaptation data for the end-to-end models. In this paper we show that RNN-transducer models can be effectively adapted to new domains using only small amounts of textual data. By taking advantage of model's inherent structure, where the prediction network is interpreted as a language model, we can apply fast adaptation to the model. Adapting the model avoids the need for complicated decoding time fusions and external language models. Using appropriate regularization, the prediction network can be adapted to new domains while still retaining good generalization capabilities. We show with multiple ASR evaluation tasks how this method can provide relative gains of 10-45% in target task WER. We also share insights how RNN-transducer prediction network performs as a language model.