Towards Automatic Methods to Detect Errors in Transcriptions of Speech Recordings

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216305%3A26230%2F19%3APU134190" target="_blank" >RIV/00216305:26230/19:PU134190 - isvavai.cz</a>

Výsledek na webu

<a href="https://ieeexplore.ieee.org/document/8683722" target="_blank" >https://ieeexplore.ieee.org/document/8683722</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1109/ICASSP.2019.8683722" target="_blank" >10.1109/ICASSP.2019.8683722</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Towards Automatic Methods to Detect Errors in Transcriptions of Speech Recordings

Popis výsledku v původním jazyce

This work explores different methods to detect errors in transcriptions of speech recordings. We artificially corrupt well transcribed speech transcriptions with three types of errors: substitution, insertion and deletion on TIMIT phonemic transcriptions and WSJ word transcriptions. First, we use Bayesian model selection method by comparing the log-likelihoods from alignment and phone recognizer, a final score is computed to make decision. In this method, we consider two models, Bayesian Hidden Markov Model (HMM) and a Variational Auto-Encoder (VAE) combined with a HMM. Alternately, we build a biased ASR system with language models trained on individual transcriptions, detection decision is based on Levenshtein distance (LD) between transcription and oracle path from decoded lattice. We evaluate the methods of detecting errors in corrupted TIMIT transcription, the best result (either using model selection with VAE model or biased ASR) achieves 7% equal error rate on the Detection Error Tradeoff (DET) curve; we also evaluate the methods of detecting errors in corrupted WSJ transcriptions, and the best result (using biased ASR) achieves 3% equal error rate.

Název v anglickém jazyce

Towards Automatic Methods to Detect Errors in Transcriptions of Speech Recordings

Popis výsledku anglicky

This work explores different methods to detect errors in transcriptions of speech recordings. We artificially corrupt well transcribed speech transcriptions with three types of errors: substitution, insertion and deletion on TIMIT phonemic transcriptions and WSJ word transcriptions. First, we use Bayesian model selection method by comparing the log-likelihoods from alignment and phone recognizer, a final score is computed to make decision. In this method, we consider two models, Bayesian Hidden Markov Model (HMM) and a Variational Auto-Encoder (VAE) combined with a HMM. Alternately, we build a biased ASR system with language models trained on individual transcriptions, detection decision is based on Levenshtein distance (LD) between transcription and oracle path from decoded lattice. We evaluate the methods of detecting errors in corrupted TIMIT transcription, the best result (either using model selection with VAE model or biased ASR) achieves 7% equal error rate on the Detection Error Tradeoff (DET) curve; we also evaluate the methods of detecting errors in corrupted WSJ transcriptions, and the best result (using biased ASR) achieves 3% equal error rate.

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/LQ1602" target="_blank" >LQ1602: IT4Innovations excellence in science</a><br>

Návaznosti

P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)<br>S - Specificky vyzkum na vysokych skolach

Rok uplatnění

2019

Kód důvěrnosti údajů

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

Název statě ve sborníku

Proceedings of ICASSP

ISBN

978-1-5386-4658-8

ISSN

—

e-ISSN

—

Počet stran výsledku

5

Strana od-do

3747-3751

Název nakladatele

IEEE Signal Processing Society

Místo vydání

Brighton

Místo konání akce

Brighton

Datum konání akce

12. 5. 2019

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

000482554003194