Improving Self-supervised Pre-training using Accent-Specific Codebooks

TL;DR

This paper introduces a novel accent-aware adaptation method for self-supervised speech recognition that uses trainable accent-specific codebooks, significantly improving performance across diverse English accents.

Contribution

It presents a new trainable codebook approach for accent adaptation in self-supervised ASR, enhancing accent invariance during pre-training and fine-tuning.

Findings

Outperforms existing accent-adaptation methods on Mozilla Common Voice

Achieves up to 9% relative reduction in word error rate

Effective on both seen and unseen accents

Abstract

Speech accents present a serious challenge to the performance of state-of-the-art end-to-end Automatic Speech Recognition (ASR) systems. Even with self-supervised learning and pre-training of ASR models, accent invariance is seldom achieved. In this work, we propose an accent-aware adaptation technique for self-supervised learning that introduces a trainable set of accent-specific codebooks to the self-supervised architecture. These learnable codebooks enable the model to capture accent specific information during pre-training, that is further refined during ASR finetuning. On the Mozilla Common Voice dataset, our proposed approach outperforms all other accent-adaptation approaches on both seen and unseen English accents, with up to 9% relative reduction in word error rate (WER).