@inproceedings{mccarthy-etal-2023-long,
title = "Long-Form Speech Translation through Segmentation with Finite-State Decoding Constraints on Large Language Models",
author = "McCarthy, Arya and
Zhang, Hao and
Kumar, Shankar and
Stahlberg, Felix and
Wu, Ke",
editor = "Bouamor, Houda and
Pino, Juan and
Bali, Kalika",
booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2023",
month = dec,
year = "2023",
address = "Singapore",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.findings-emnlp.19",
doi = "10.18653/v1/2023.findings-emnlp.19",
pages = "247--257",
abstract = "One challenge in speech translation is that plenty of spoken content is long-form, but short units are necessary for obtaining high-quality translations. To address this mismatch, we adapt large language models (LLMs) to split long ASR transcripts into segments that can be independently translated so as to maximize the overall translation quality. We overcome the tendency of hallucination in LLMs by incorporating finite-state constraints during decoding; these eliminate invalid outputs without requiring additional training. We discover that LLMs are adaptable to transcripts containing ASR errors through prompt-tuning or fine-tuning. Relative to a state-of-the-art automatic punctuation baseline, our best LLM improves the average BLEU by 2.9 points for English{--}German, English{--}Spanish, and English{--}Arabic TED talk translation in 9 test sets, just by improving segmentation.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="mccarthy-etal-2023-long">
<titleInfo>
<title>Long-Form Speech Translation through Segmentation with Finite-State Decoding Constraints on Large Language Models</title>
</titleInfo>
<name type="personal">
<namePart type="given">Arya</namePart>
<namePart type="family">McCarthy</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hao</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Shankar</namePart>
<namePart type="family">Kumar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Felix</namePart>
<namePart type="family">Stahlberg</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ke</namePart>
<namePart type="family">Wu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2023-12</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: EMNLP 2023</title>
</titleInfo>
<name type="personal">
<namePart type="given">Houda</namePart>
<namePart type="family">Bouamor</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Juan</namePart>
<namePart type="family">Pino</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Kalika</namePart>
<namePart type="family">Bali</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Singapore</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>One challenge in speech translation is that plenty of spoken content is long-form, but short units are necessary for obtaining high-quality translations. To address this mismatch, we adapt large language models (LLMs) to split long ASR transcripts into segments that can be independently translated so as to maximize the overall translation quality. We overcome the tendency of hallucination in LLMs by incorporating finite-state constraints during decoding; these eliminate invalid outputs without requiring additional training. We discover that LLMs are adaptable to transcripts containing ASR errors through prompt-tuning or fine-tuning. Relative to a state-of-the-art automatic punctuation baseline, our best LLM improves the average BLEU by 2.9 points for English–German, English–Spanish, and English–Arabic TED talk translation in 9 test sets, just by improving segmentation.</abstract>
<identifier type="citekey">mccarthy-etal-2023-long</identifier>
<identifier type="doi">10.18653/v1/2023.findings-emnlp.19</identifier>
<location>
<url>https://aclanthology.org/2023.findings-emnlp.19</url>
</location>
<part>
<date>2023-12</date>
<extent unit="page">
<start>247</start>
<end>257</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Long-Form Speech Translation through Segmentation with Finite-State Decoding Constraints on Large Language Models
%A McCarthy, Arya
%A Zhang, Hao
%A Kumar, Shankar
%A Stahlberg, Felix
%A Wu, Ke
%Y Bouamor, Houda
%Y Pino, Juan
%Y Bali, Kalika
%S Findings of the Association for Computational Linguistics: EMNLP 2023
%D 2023
%8 December
%I Association for Computational Linguistics
%C Singapore
%F mccarthy-etal-2023-long
%X One challenge in speech translation is that plenty of spoken content is long-form, but short units are necessary for obtaining high-quality translations. To address this mismatch, we adapt large language models (LLMs) to split long ASR transcripts into segments that can be independently translated so as to maximize the overall translation quality. We overcome the tendency of hallucination in LLMs by incorporating finite-state constraints during decoding; these eliminate invalid outputs without requiring additional training. We discover that LLMs are adaptable to transcripts containing ASR errors through prompt-tuning or fine-tuning. Relative to a state-of-the-art automatic punctuation baseline, our best LLM improves the average BLEU by 2.9 points for English–German, English–Spanish, and English–Arabic TED talk translation in 9 test sets, just by improving segmentation.
%R 10.18653/v1/2023.findings-emnlp.19
%U https://aclanthology.org/2023.findings-emnlp.19
%U https://doi.org/10.18653/v1/2023.findings-emnlp.19
%P 247-257
Markdown (Informal)
[Long-Form Speech Translation through Segmentation with Finite-State Decoding Constraints on Large Language Models](https://aclanthology.org/2023.findings-emnlp.19) (McCarthy et al., Findings 2023)
ACL