@inproceedings{lei-etal-2023-ckdst,
title = "{CKDST}: Comprehensively and Effectively Distill Knowledge from Machine Translation to End-to-End Speech Translation",
author = "Lei, Yikun and
Xue, Zhengshan and
Zhao, Xiaohu and
Sun, Haoran and
Zhu, Shaolin and
Lin, Xiaodong and
Xiong, Deyi",
editor = "Rogers, Anna and
Boyd-Graber, Jordan and
Okazaki, Naoaki",
booktitle = "Findings of the Association for Computational Linguistics: ACL 2023",
month = jul,
year = "2023",
address = "Toronto, Canada",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2023.findings-acl.195",
doi = "10.18653/v1/2023.findings-acl.195",
pages = "3123--3137",
abstract = "Distilling knowledge from a high-resource task, e.g., machine translation, is an effective way to alleviate the data scarcity problem of end-to-end speech translation. However, previous works simply use the classical knowledge distillation that does not allow for adequate transfer of knowledge from machine translation. In this paper, we propose a comprehensive knowledge distillation framework for speech translation, CKDST, which is capable of comprehensively and effectively distilling knowledge from machine translation to speech translation from two perspectives: cross-modal contrastive representation distillation and simultaneous decoupled knowledge distillation. In the former, we leverage a contrastive learning objective to optmize the mutual information between speech and text representations for representation distillation in the encoder. In the later, we decouple the non-target class knowledge from target class knowledge for logits distillation in the decoder. Experiments on the MuST-C benchmark dataset demonstrate that our CKDST substantially improves the baseline by 1.2 BLEU on average in all translation directions, and outperforms previous state-of-the-art end-to-end and cascaded speech translation models.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="lei-etal-2023-ckdst">
<titleInfo>
<title>CKDST: Comprehensively and Effectively Distill Knowledge from Machine Translation to End-to-End Speech Translation</title>
</titleInfo>
<name type="personal">
<namePart type="given">Yikun</namePart>
<namePart type="family">Lei</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Zhengshan</namePart>
<namePart type="family">Xue</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xiaohu</namePart>
<namePart type="family">Zhao</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Haoran</namePart>
<namePart type="family">Sun</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Shaolin</namePart>
<namePart type="family">Zhu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xiaodong</namePart>
<namePart type="family">Lin</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Deyi</namePart>
<namePart type="family">Xiong</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2023-07</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: ACL 2023</title>
</titleInfo>
<name type="personal">
<namePart type="given">Anna</namePart>
<namePart type="family">Rogers</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jordan</namePart>
<namePart type="family">Boyd-Graber</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Naoaki</namePart>
<namePart type="family">Okazaki</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Toronto, Canada</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Distilling knowledge from a high-resource task, e.g., machine translation, is an effective way to alleviate the data scarcity problem of end-to-end speech translation. However, previous works simply use the classical knowledge distillation that does not allow for adequate transfer of knowledge from machine translation. In this paper, we propose a comprehensive knowledge distillation framework for speech translation, CKDST, which is capable of comprehensively and effectively distilling knowledge from machine translation to speech translation from two perspectives: cross-modal contrastive representation distillation and simultaneous decoupled knowledge distillation. In the former, we leverage a contrastive learning objective to optmize the mutual information between speech and text representations for representation distillation in the encoder. In the later, we decouple the non-target class knowledge from target class knowledge for logits distillation in the decoder. Experiments on the MuST-C benchmark dataset demonstrate that our CKDST substantially improves the baseline by 1.2 BLEU on average in all translation directions, and outperforms previous state-of-the-art end-to-end and cascaded speech translation models.</abstract>
<identifier type="citekey">lei-etal-2023-ckdst</identifier>
<identifier type="doi">10.18653/v1/2023.findings-acl.195</identifier>
<location>
<url>https://aclanthology.org/2023.findings-acl.195</url>
</location>
<part>
<date>2023-07</date>
<extent unit="page">
<start>3123</start>
<end>3137</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T CKDST: Comprehensively and Effectively Distill Knowledge from Machine Translation to End-to-End Speech Translation
%A Lei, Yikun
%A Xue, Zhengshan
%A Zhao, Xiaohu
%A Sun, Haoran
%A Zhu, Shaolin
%A Lin, Xiaodong
%A Xiong, Deyi
%Y Rogers, Anna
%Y Boyd-Graber, Jordan
%Y Okazaki, Naoaki
%S Findings of the Association for Computational Linguistics: ACL 2023
%D 2023
%8 July
%I Association for Computational Linguistics
%C Toronto, Canada
%F lei-etal-2023-ckdst
%X Distilling knowledge from a high-resource task, e.g., machine translation, is an effective way to alleviate the data scarcity problem of end-to-end speech translation. However, previous works simply use the classical knowledge distillation that does not allow for adequate transfer of knowledge from machine translation. In this paper, we propose a comprehensive knowledge distillation framework for speech translation, CKDST, which is capable of comprehensively and effectively distilling knowledge from machine translation to speech translation from two perspectives: cross-modal contrastive representation distillation and simultaneous decoupled knowledge distillation. In the former, we leverage a contrastive learning objective to optmize the mutual information between speech and text representations for representation distillation in the encoder. In the later, we decouple the non-target class knowledge from target class knowledge for logits distillation in the decoder. Experiments on the MuST-C benchmark dataset demonstrate that our CKDST substantially improves the baseline by 1.2 BLEU on average in all translation directions, and outperforms previous state-of-the-art end-to-end and cascaded speech translation models.
%R 10.18653/v1/2023.findings-acl.195
%U https://aclanthology.org/2023.findings-acl.195
%U https://doi.org/10.18653/v1/2023.findings-acl.195
%P 3123-3137
Markdown (Informal)
[CKDST: Comprehensively and Effectively Distill Knowledge from Machine Translation to End-to-End Speech Translation](https://aclanthology.org/2023.findings-acl.195) (Lei et al., Findings 2023)
ACL