Bayesian joint-sequence models for grapheme-to-phoneme conversion

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216305%3A26230%2F17%3APU126426" target="_blank" >RIV/00216305:26230/17:PU126426 - isvavai.cz</a>

Výsledek na webu

<a href="https://www.fit.vut.cz/research/publication/11469/" target="_blank" >https://www.fit.vut.cz/research/publication/11469/</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1109/ICASSP.2017.7952674" target="_blank" >10.1109/ICASSP.2017.7952674</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Bayesian joint-sequence models for grapheme-to-phoneme conversion

Popis výsledku v původním jazyce

We describe a fully Bayesian approach to grapheme-to-phoneme conversion based on the joint-sequence model (JSM). Usually, standard smoothed n-gram language models (LM, e.g. Kneser-Ney) are used with JSMs to model graphone sequences (joint graphemephoneme pairs). However, we take a Bayesian approach using a hierarchical Pitman-Yor-Process LM. This provides an elegant alternative to using smoothing techniques to avoid over-training. No held-out sets and complex parameter tuning is necessary, and several convergence problems encountered in the discounted Expectation- Maximization (as used in the smoothed JSMs) are avoided. Every step is modeled by weighted finite state transducers and implemented with standard operations from the OpenFST toolkit. We evaluate our model on a standard data set (CMUdict), where it gives comparable results to the previously reported smoothed JSMs in terms of phoneme-error rate while requiring a much smaller training/ testing time. Most importantly, our model can be used in a Bayesian framework and for (partly) un-supervised training.

Název v anglickém jazyce

Bayesian joint-sequence models for grapheme-to-phoneme conversion

Popis výsledku anglicky

We describe a fully Bayesian approach to grapheme-to-phoneme conversion based on the joint-sequence model (JSM). Usually, standard smoothed n-gram language models (LM, e.g. Kneser-Ney) are used with JSMs to model graphone sequences (joint graphemephoneme pairs). However, we take a Bayesian approach using a hierarchical Pitman-Yor-Process LM. This provides an elegant alternative to using smoothing techniques to avoid over-training. No held-out sets and complex parameter tuning is necessary, and several convergence problems encountered in the discounted Expectation- Maximization (as used in the smoothed JSMs) are avoided. Every step is modeled by weighted finite state transducers and implemented with standard operations from the OpenFST toolkit. We evaluate our model on a standard data set (CMUdict), where it gives comparable results to the previously reported smoothed JSMs in terms of phoneme-error rate while requiring a much smaller training/ testing time. Most importantly, our model can be used in a Bayesian framework and for (partly) un-supervised training.

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

—

Návaznosti

N - Vyzkumna aktivita podporovana z neverejnych zdroju

Rok uplatnění

2017

Kód důvěrnosti údajů

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

Název statě ve sborníku

Proceedings of ICASSP 2017

ISBN

978-1-5090-4117-6

ISSN

—

e-ISSN

—

Počet stran výsledku

5

Strana od-do

2836-2840

Název nakladatele

IEEE Signal Processing Society

Místo vydání

New Orleans

Místo konání akce

New Orleans, USA

Datum konání akce

5. 3. 2017

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

000414286203002