@inproceedings{salesky-black-2020-phone,
title = "Phone Features Improve Speech Translation",
author = "Salesky, Elizabeth and
Black, Alan W",
booktitle = "Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics",
month = jul,
year = "2020",
address = "Online",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2020.acl-main.217",
doi = "10.18653/v1/2020.acl-main.217",
pages = "2388--2397",
abstract = "End-to-end models for speech translation (ST) more tightly couple speech recognition (ASR) and machine translation (MT) than a traditional cascade of separate ASR and MT models, with simpler model architectures and the potential for reduced error propagation. Their performance is often assumed to be superior, though in many conditions this is not yet the case. We compare cascaded and end-to-end models across high, medium, and low-resource conditions, and show that cascades remain stronger baselines. Further, we introduce two methods to incorporate phone features into ST models. We show that these features improve both architectures, closing the gap between end-to-end models and cascades, and outperforming previous academic work {--} by up to 9 BLEU on our low-resource setting.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="salesky-black-2020-phone">
<titleInfo>
<title>Phone Features Improve Speech Translation</title>
</titleInfo>
<name type="personal">
<namePart type="given">Elizabeth</namePart>
<namePart type="family">Salesky</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Alan</namePart>
<namePart type="given">W</namePart>
<namePart type="family">Black</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2020-jul</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics</title>
</titleInfo>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Online</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>End-to-end models for speech translation (ST) more tightly couple speech recognition (ASR) and machine translation (MT) than a traditional cascade of separate ASR and MT models, with simpler model architectures and the potential for reduced error propagation. Their performance is often assumed to be superior, though in many conditions this is not yet the case. We compare cascaded and end-to-end models across high, medium, and low-resource conditions, and show that cascades remain stronger baselines. Further, we introduce two methods to incorporate phone features into ST models. We show that these features improve both architectures, closing the gap between end-to-end models and cascades, and outperforming previous academic work – by up to 9 BLEU on our low-resource setting.</abstract>
<identifier type="citekey">salesky-black-2020-phone</identifier>
<identifier type="doi">10.18653/v1/2020.acl-main.217</identifier>
<location>
<url>https://aclanthology.org/2020.acl-main.217</url>
</location>
<part>
<date>2020-jul</date>
<extent unit="page">
<start>2388</start>
<end>2397</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Phone Features Improve Speech Translation
%A Salesky, Elizabeth
%A Black, Alan W.
%S Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics
%D 2020
%8 jul
%I Association for Computational Linguistics
%C Online
%F salesky-black-2020-phone
%X End-to-end models for speech translation (ST) more tightly couple speech recognition (ASR) and machine translation (MT) than a traditional cascade of separate ASR and MT models, with simpler model architectures and the potential for reduced error propagation. Their performance is often assumed to be superior, though in many conditions this is not yet the case. We compare cascaded and end-to-end models across high, medium, and low-resource conditions, and show that cascades remain stronger baselines. Further, we introduce two methods to incorporate phone features into ST models. We show that these features improve both architectures, closing the gap between end-to-end models and cascades, and outperforming previous academic work – by up to 9 BLEU on our low-resource setting.
%R 10.18653/v1/2020.acl-main.217
%U https://aclanthology.org/2020.acl-main.217
%U https://doi.org/10.18653/v1/2020.acl-main.217
%P 2388-2397
Markdown (Informal)
[Phone Features Improve Speech Translation](https://aclanthology.org/2020.acl-main.217) (Salesky & Black, ACL 2020)
ACL
- Elizabeth Salesky and Alan W Black. 2020. Phone Features Improve Speech Translation. In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics, pages 2388–2397, Online. Association for Computational Linguistics.