Text Augmentation for Language Models in High Error Recognition Scenario

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216305%3A26230%2F21%3APU142964" target="_blank" >RIV/00216305:26230/21:PU142964 - isvavai.cz</a>

Výsledek na webu

<a href="https://www.isca-speech.org/archive/interspeech_2021/benes21_interspeech.html" target="_blank" >https://www.isca-speech.org/archive/interspeech_2021/benes21_interspeech.html</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.21437/Interspeech.2021-627" target="_blank" >10.21437/Interspeech.2021-627</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Text Augmentation for Language Models in High Error Recognition Scenario

Popis výsledku v původním jazyce

In this paper, we explore several data augmentation strategies for training of language models for speech recognition. We compare augmentation based on global error statistics with one based on unigram statistics of ASR errors and with labelsmoothing and its sampled variant. Additionally, we investigate the stability and the predictive power of perplexity estimated on augmented data. Despite being trivial, augmentation driven by global substitution, deletion and insertion rates achieves the best rescoring results. On the other hand, even though the associated perplexity measure is stable, it gives no better prediction of the final error rate than the vanilla one. Our best augmentation scheme increases the WER improvement from second-pass rescoring from 1.1% to 1.9% absolute on the CHiMe-6 challenge.

Název v anglickém jazyce

Text Augmentation for Language Models in High Error Recognition Scenario

Popis výsledku anglicky

In this paper, we explore several data augmentation strategies for training of language models for speech recognition. We compare augmentation based on global error statistics with one based on unigram statistics of ASR errors and with labelsmoothing and its sampled variant. Additionally, we investigate the stability and the predictive power of perplexity estimated on augmented data. Despite being trivial, augmentation driven by global substitution, deletion and insertion rates achieves the best rescoring results. On the other hand, even though the associated perplexity measure is stable, it gives no better prediction of the final error rate than the vanilla one. Our best augmentation scheme increases the WER improvement from second-pass rescoring from 1.1% to 1.9% absolute on the CHiMe-6 challenge.

Druh

D - Stať ve sborníku

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

<a href="/cs/project/GX19-26934X" target="_blank" >GX19-26934X: Neuronové reprezentace v multimodálním a mnohojazyčném modelování</a><br>

Návaznosti

P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)

Rok uplatnění

2021

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH

ISBN

—

ISSN

1990-9772

e-ISSN

—

Počet stran výsledku

5

Strana od-do

1872-1876

Název nakladatele

International Speech Communication Association

Místo vydání

Brno

Místo konání akce

Brno

Datum konání akce

30. 8. 2021

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

000841879501198