@inproceedings{havard-etal-2025-speech,
title = "Speech Technologies with Fieldwork Recordings: the Case of {H}aitian {C}reole",
author = "Havard, William N. and
Govain, Renauld and
Lecouteux, Benjamin and
Schang, Emmanuel",
editor = "Lachler, Jordan and
Agyapong, Godfred and
Arppe, Antti and
Moeller, Sarah and
Chaudhary, Aditi and
Rijhwani, Shruti and
Rosenblum, Daisy",
booktitle = "Proceedings of the Eight Workshop on the Use of Computational Methods in the Study of Endangered Languages",
month = mar,
year = "2025",
address = "Honolulu, Hawaii, USA",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2025.computel-main.5/",
pages = "40--46",
abstract = "We use 40-year-old digitalised tape-recorded fieldwork data in Haitian Creole to train a native self-supervised learning (SSL) model of speech representation (WAV2VEC2). We also use a continued pre-training approach on pre-trained SSL models of two foreign languages: the lexifier language {--} French {--} and an unrelated language {--} English. We compare the performances of these three SSL models, and of two other foreign SSL models directly finetuned, on an ASR task, where all five models are fine-tuned on transcribed fieldwork recordings in Haitian Creole. Our results show the best-performing model is the one trained using a continued pre-training approach on the lexifier language, followed by the native model. We conclude that the `mobilising the archive'-approach advocated by (Bird, 2020) is a promising way forward to design speech technologies for new languages."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="havard-etal-2025-speech">
<titleInfo>
<title>Speech Technologies with Fieldwork Recordings: the Case of Haitian Creole</title>
</titleInfo>
<name type="personal">
<namePart type="given">William</namePart>
<namePart type="given">N</namePart>
<namePart type="family">Havard</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Renauld</namePart>
<namePart type="family">Govain</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Benjamin</namePart>
<namePart type="family">Lecouteux</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Emmanuel</namePart>
<namePart type="family">Schang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2025-03</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the Eight Workshop on the Use of Computational Methods in the Study of Endangered Languages</title>
</titleInfo>
<name type="personal">
<namePart type="given">Jordan</namePart>
<namePart type="family">Lachler</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Godfred</namePart>
<namePart type="family">Agyapong</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Antti</namePart>
<namePart type="family">Arppe</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sarah</namePart>
<namePart type="family">Moeller</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Aditi</namePart>
<namePart type="family">Chaudhary</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Shruti</namePart>
<namePart type="family">Rijhwani</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Daisy</namePart>
<namePart type="family">Rosenblum</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Honolulu, Hawaii, USA</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>We use 40-year-old digitalised tape-recorded fieldwork data in Haitian Creole to train a native self-supervised learning (SSL) model of speech representation (WAV2VEC2). We also use a continued pre-training approach on pre-trained SSL models of two foreign languages: the lexifier language – French – and an unrelated language – English. We compare the performances of these three SSL models, and of two other foreign SSL models directly finetuned, on an ASR task, where all five models are fine-tuned on transcribed fieldwork recordings in Haitian Creole. Our results show the best-performing model is the one trained using a continued pre-training approach on the lexifier language, followed by the native model. We conclude that the ‘mobilising the archive’-approach advocated by (Bird, 2020) is a promising way forward to design speech technologies for new languages.</abstract>
<identifier type="citekey">havard-etal-2025-speech</identifier>
<location>
<url>https://aclanthology.org/2025.computel-main.5/</url>
</location>
<part>
<date>2025-03</date>
<extent unit="page">
<start>40</start>
<end>46</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Speech Technologies with Fieldwork Recordings: the Case of Haitian Creole
%A Havard, William N.
%A Govain, Renauld
%A Lecouteux, Benjamin
%A Schang, Emmanuel
%Y Lachler, Jordan
%Y Agyapong, Godfred
%Y Arppe, Antti
%Y Moeller, Sarah
%Y Chaudhary, Aditi
%Y Rijhwani, Shruti
%Y Rosenblum, Daisy
%S Proceedings of the Eight Workshop on the Use of Computational Methods in the Study of Endangered Languages
%D 2025
%8 March
%I Association for Computational Linguistics
%C Honolulu, Hawaii, USA
%F havard-etal-2025-speech
%X We use 40-year-old digitalised tape-recorded fieldwork data in Haitian Creole to train a native self-supervised learning (SSL) model of speech representation (WAV2VEC2). We also use a continued pre-training approach on pre-trained SSL models of two foreign languages: the lexifier language – French – and an unrelated language – English. We compare the performances of these three SSL models, and of two other foreign SSL models directly finetuned, on an ASR task, where all five models are fine-tuned on transcribed fieldwork recordings in Haitian Creole. Our results show the best-performing model is the one trained using a continued pre-training approach on the lexifier language, followed by the native model. We conclude that the ‘mobilising the archive’-approach advocated by (Bird, 2020) is a promising way forward to design speech technologies for new languages.
%U https://aclanthology.org/2025.computel-main.5/
%P 40-46
Markdown (Informal)
[Speech Technologies with Fieldwork Recordings: the Case of Haitian Creole](https://aclanthology.org/2025.computel-main.5/) (Havard et al., ComputEL 2025)
ACL