Customising Czech phonetic alignment using HuBERT and manual segmentation

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216208%3A11210%2F25%3A10508903" target="_blank" >RIV/00216208:11210/25:10508903 - isvavai.cz</a>

Výsledek na webu

<a href="https://verso.is.cuni.cz/pub/verso.fpl?fname=obd_publikace_handle&handle=M1bzW3Xugi" target="_blank" >https://verso.is.cuni.cz/pub/verso.fpl?fname=obd_publikace_handle&handle=M1bzW3Xugi</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.14712/24646830.2025.20" target="_blank" >10.14712/24646830.2025.20</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Customising Czech phonetic alignment using HuBERT and manual segmentation

Popis výsledku v původním jazyce

This paper presents Prak, a forced alignment tool developed for Czech, with a focus on transparent modular design and phonetic accuracy. In addition to a rule-based pronunciation module and exception handling, Prak introduces a novel application of non-deterministic, backward-processing FSTs to model complex regressive assimilation processes in Czech consonant clusters. We further describe the integration of a HuBERT-based transformer model and training including extensive manually time-aligned data to enhance phone classification accuracy while maintaining ease of installation and use. Evaluation against a manually aligned test corpus demonstrates that the enhanced model significantly outperforms both our earlier Prak-CV model and the long-established previous forced alignment baseline. The new model reduces major boundary errors and mismatches, bringing alignment accuracy closer to manual phonetic segmentation standards for Czech. We emphasize both methodological transparency and practical usability, aiming to support phoneticians working with Czech as well as developers interested in extending the tool for other languages.

Název v anglickém jazyce

Customising Czech phonetic alignment using HuBERT and manual segmentation

Popis výsledku anglicky

This paper presents Prak, a forced alignment tool developed for Czech, with a focus on transparent modular design and phonetic accuracy. In addition to a rule-based pronunciation module and exception handling, Prak introduces a novel application of non-deterministic, backward-processing FSTs to model complex regressive assimilation processes in Czech consonant clusters. We further describe the integration of a HuBERT-based transformer model and training including extensive manually time-aligned data to enhance phone classification accuracy while maintaining ease of installation and use. Evaluation against a manually aligned test corpus demonstrates that the enhanced model significantly outperforms both our earlier Prak-CV model and the long-established previous forced alignment baseline. The new model reduces major boundary errors and mismatches, bringing alignment accuracy closer to manual phonetic segmentation standards for Czech. We emphasize both methodological transparency and practical usability, aiming to support phoneticians working with Czech as well as developers interested in extending the tool for other languages.

Druh

J_ost - Ostatní články v recenzovaných periodicích

CEP obor

—

OECD FORD obor

60203 - Linguistics

Projekt

—

Návaznosti

I - Institucionalni podpora na dlouhodoby koncepcni rozvoj vyzkumne organizace

Rok uplatnění

2025

Kód důvěrnosti údajů

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

Název periodika

Acta Universitatis Carolinae. Philologica

ISSN

0567-8269

e-ISSN

2464-6830

Svazek periodika

2025

Číslo periodika v rámci svazku

3

Stát vydavatele periodika

CZ - Česká republika

Počet stran výsledku

18

Strana od-do

43-60

Kód UT WoS článku

—

EID výsledku v databázi Scopus

—