Understanding Shared Speech-Text Representations
Abstract
Recently, a number of approaches to train speech models by incorpo-rating text into end-to-end models have been developed, with Mae-stro advancing state-of-the-art automatic speech recognition (ASR)and Speech Translation (ST) performance. In this paper, we expandour understanding of the resulting shared speech-text representationswith two types of analyses. First we examine the limits of speech-free domain adaptation, finding that a corpus-specific duration modelfor speech-text alignment is the most important component for learn-ing a shared speech-text representation. Second, we inspect the sim-ilarities between activations of unimodal (speech or text) encodersas compared to the activations of a shared encoder. We find that theshared encoder learns a more compact and overlapping speech-textrepresentation than the uni-modal encoders. We hypothesize that thispartially explains the effectiveness of the Maestro shared speech-textrepresentations.
Cite
@article{arxiv.2304.14514,
title = {Understanding Shared Speech-Text Representations},
author = {Gary Wang and Kyle Kastner and Ankur Bapna and Zhehuai Chen and Andrew Rosenberg and Bhuvana Ramabhadran and Yu Zhang},
journal= {arXiv preprint arXiv:2304.14514},
year = {2023}
}
Comments
Accepted at ICASSP 2023, camera ready