BUT Systems for IWSLT 2023 Marathi - Hindi Low Resource Speech Translation Task

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216305%3A26230%2F23%3APU149425" target="_blank" >RIV/00216305:26230/23:PU149425 - isvavai.cz</a>

Výsledek na webu

<a href="https://aclanthology.org/2023.iwslt-1.19.pdf" target="_blank" >https://aclanthology.org/2023.iwslt-1.19.pdf</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.18653/v1/2023.iwslt-1.19" target="_blank" >10.18653/v1/2023.iwslt-1.19</a>

Jazyk výsledku

angličtina

Název v původním jazyce

BUT Systems for IWSLT 2023 Marathi - Hindi Low Resource Speech Translation Task

Popis výsledku v původním jazyce

This paper describes the systems submitted for Marathi to Hindi low-resource speech translation task. Our primary submission is based on an end-to-end direct speech translation system, whereas the contrastive one is a cascaded system. The backbone of both the systems is a Hindi-Marathi bilingual ASR system trained on 2790 hours of imperfect transcribed speech. The end-to-end speech translation system was directly initialized from the ASR, and then finetuned for direct speech translation with an auxiliary CTC loss for translation. The MT model for the cascaded system is initialized from a cross-lingual language model, which was then fine-tuned using 1.6 M parallel sentences. All our systems were trained from scratch on publicly available datasets. In the end, we use a language model to re-score the n-best hypotheses. Our primary submission achieved 30.5 and 39.6 BLEU whereas the contrastive system obtained 21.7 and 28.6 BLEU on official dev and test sets respectively. The paper also presents the analysis on several experiments that were conducted and outlines the strategies for improving speech translation in low-resource scenarios.

Název v anglickém jazyce

BUT Systems for IWSLT 2023 Marathi - Hindi Low Resource Speech Translation Task

Popis výsledku anglicky

This paper describes the systems submitted for Marathi to Hindi low-resource speech translation task. Our primary submission is based on an end-to-end direct speech translation system, whereas the contrastive one is a cascaded system. The backbone of both the systems is a Hindi-Marathi bilingual ASR system trained on 2790 hours of imperfect transcribed speech. The end-to-end speech translation system was directly initialized from the ASR, and then finetuned for direct speech translation with an auxiliary CTC loss for translation. The MT model for the cascaded system is initialized from a cross-lingual language model, which was then fine-tuned using 1.6 M parallel sentences. All our systems were trained from scratch on publicly available datasets. In the end, we use a language model to re-score the n-best hypotheses. Our primary submission achieved 30.5 and 39.6 BLEU whereas the contrastive system obtained 21.7 and 28.6 BLEU on official dev and test sets respectively. The paper also presents the analysis on several experiments that were conducted and outlines the strategies for improving speech translation in low-resource scenarios.

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

Výsledek vznikl pri realizaci vícero projektů. Více informací v záložce Projekty.

Návaznosti

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

Rok uplatnění

2023

Kód důvěrnosti údajů

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

Název statě ve sborníku

20th International Conference on Spoken Language Translation, IWSLT 2023 - Proceedings of the Conference

ISBN

978-1-959429-84-5

ISSN

—

e-ISSN

—

Počet stran výsledku

8

Strana od-do

227-234

Název nakladatele

Association for Computational Linguistics

Místo vydání

Toronto (in-person and online)

Místo konání akce

Toronto

Datum konání akce

9. 7. 2023

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

—