Optimizing Byte-level Representation for End-to-end ASR

TL;DR

This paper introduces a method to optimize byte-level representations for end-to-end ASR using auto-encoder and vector quantization, resulting in improved accuracy over standard UTF-8 encoding in multilingual speech recognition.

Contribution

It presents a novel framework that optimizes byte-level representations specifically for ASR, incorporating multimodal information and error correction, outperforming traditional UTF-8 encoding.

Findings

Achieved 5% relative error rate reduction in English/Mandarin ASR

Demonstrated effectiveness of optimized byte representations over UTF-8

Proposed a versatile framework integrating multiple modalities

Abstract

We propose a novel approach to optimizing a byte-level representation for end-to-end automatic speech recognition (ASR). Byte-level representation is often used by large scale multilingual ASR systems when the character set of the supported languages is large. The compactness and universality of byte-level representation allow the ASR models to use smaller output vocabularies and therefore, provide more flexibility. UTF-8 is a commonly used byte-level representation for multilingual ASR, but it is not designed to optimize machine learning tasks directly. By using auto-encoder and vector quantization, we show that we can optimize a byte-level representation for ASR and achieve better accuracy. Our proposed framework can incorporate information from different modalities, and provides an error correction mechanism. In an English/Mandarin dictation task, we show that a bilingual ASR model built with this approach can outperform UTF-8 representation by 5% relative in error rate.