Advancing Multi-Accented LSTM-CTC Speech Recognition using a Domain Specific Student-Teacher Learning Paradigm

TL;DR

This paper introduces a domain-specific student-teacher learning paradigm for multi-accent speech recognition using LSTM-CTC models, significantly improving accuracy across diverse accents by leveraging aligned accent-specific teachers and knowledge distillation.

Contribution

It proposes a novel multi-accent learning framework with aligned accent-specific teacher models and a student model, achieving substantial CER reduction and effective accent adaptation.

Findings

20.1% relative CER reduction with the proposed method

Aligned accent-specific models improve recognition accuracy

Knowledge distillation enhances accent adaptation performance

Abstract

Non-native speech causes automatic speech recognition systems to degrade in performance. Past strategies to address this challenge have considered model adaptation, accent classification with a model selection, alternate pronunciation lexicon, etc. In this study, we consider a recurrent neural network (RNN) with connectionist temporal classification (CTC) cost function trained on multi-accent English data including US (Native), Indian and Hispanic accents. We exploit dark knowledge from a model trained with the multi-accent data to train student models under the guidance of both a teacher model and CTC cost of target transcription. We show that transferring knowledge from a single RNN-CTC trained model toward a student model, yields better performance than the stand-alone teacher model. Since the outputs of different trained CTC models are not necessarily aligned, it is not possible to simply use an ensemble of CTC teacher models. To address this problem, we train accent specific models under the guidance of a single multi-accent teacher, which results in having multiple aligned and trained CTC models. Furthermore, we train a student model under the supervision of the accent-specific teachers, resulting in an even further complementary model, which achieves +20.1% relative Character Error Rate (CER) reduction compared to the baseline trained without any teacher. Having this effective multi-accent model, we can achieve further improvement for each accent by adapting the model to each accent. Using the accent specific model's outputs to regularize the adapting process (i.e., a knowledge distillation version of Kullback-Leibler (KL) divergence) results in even superior performance compared to the conventional approach using general teacher models.