@inproceedings{andrade-etal-2026-less,
title = "Less can be More: Towards a Parameter-Efficient Fine-Tuning of {W}av2{V}ec2 {XLSR} for Low-Resource Cape Verdean Creole {ASR}",
author = "Andrade, Mateus Neves and
Ba, Mouhamadou Lamine and
Diop, Idy and
da Veiga, Arlindo Oliveira",
editor = "Matfunjwa, Muzi and
Setaka, Mmasibidi and
Mabuya, Rooweither and
van Zaanen, Menno",
booktitle = "Proceedings of Resources for {A}frican Indigenous Languages ({RAIL}) 2026 @ {LREC} 2026",
month = may,
year = "2026",
address = "Palma, Mallorca (Spain)",
publisher = "ELRA Language Resources Association (ELRA)",
url = "https://aclanthology.org/2026.rail-1.7/",
doi = "10.63317/4b6ij7awds27",
pages = "62--71",
abstract = "Automatic Speech Recognition (ASR) for low-resource languages remains challenging due to limited annotated data and high linguistic variability. In this work, we investigate parameter-efficient fine-tuning strategies for Cape Verdean Creole ASR using the Wav2Vec 2.0 XLSR model. We evaluate the impact of structured layer freezing on model performance, training stability, and computational efficiency. Experiments conducted on a newly curated Santiago-dialect dataset show that full fine-tuning achieves the best absolute performance (WER 0.212, CER 0.120). However, several freezing configurations achieve comparable recognition performance while substantially reducing the number of trainable parameters and exhibiting more stable convergence. These results highlight a trade-off between adaptability and efficiency, showing that selective freezing can serve as an effective regularization strategy in low-resource settings. This work provides practical insights into parameter-efficient adaptation for under-resourced Creole languages."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="andrade-etal-2026-less">
<titleInfo>
<title>Less can be More: Towards a Parameter-Efficient Fine-Tuning of Wav2Vec2 XLSR for Low-Resource Cape Verdean Creole ASR</title>
</titleInfo>
<name type="personal">
<namePart type="given">Mateus</namePart>
<namePart type="given">Neves</namePart>
<namePart type="family">Andrade</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mouhamadou</namePart>
<namePart type="given">Lamine</namePart>
<namePart type="family">Ba</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Idy</namePart>
<namePart type="family">Diop</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Arlindo</namePart>
<namePart type="given">Oliveira</namePart>
<namePart type="family">da Veiga</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-05</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of Resources for African Indigenous Languages (RAIL) 2026 @ LREC 2026</title>
</titleInfo>
<name type="personal">
<namePart type="given">Muzi</namePart>
<namePart type="family">Matfunjwa</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mmasibidi</namePart>
<namePart type="family">Setaka</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Rooweither</namePart>
<namePart type="family">Mabuya</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Menno</namePart>
<namePart type="family">van Zaanen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>ELRA Language Resources Association (ELRA)</publisher>
<place>
<placeTerm type="text">Palma, Mallorca (Spain)</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>Automatic Speech Recognition (ASR) for low-resource languages remains challenging due to limited annotated data and high linguistic variability. In this work, we investigate parameter-efficient fine-tuning strategies for Cape Verdean Creole ASR using the Wav2Vec 2.0 XLSR model. We evaluate the impact of structured layer freezing on model performance, training stability, and computational efficiency. Experiments conducted on a newly curated Santiago-dialect dataset show that full fine-tuning achieves the best absolute performance (WER 0.212, CER 0.120). However, several freezing configurations achieve comparable recognition performance while substantially reducing the number of trainable parameters and exhibiting more stable convergence. These results highlight a trade-off between adaptability and efficiency, showing that selective freezing can serve as an effective regularization strategy in low-resource settings. This work provides practical insights into parameter-efficient adaptation for under-resourced Creole languages.</abstract>
<identifier type="citekey">andrade-etal-2026-less</identifier>
<identifier type="doi">10.63317/4b6ij7awds27</identifier>
<location>
<url>https://aclanthology.org/2026.rail-1.7/</url>
</location>
<part>
<date>2026-05</date>
<extent unit="page">
<start>62</start>
<end>71</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Less can be More: Towards a Parameter-Efficient Fine-Tuning of Wav2Vec2 XLSR for Low-Resource Cape Verdean Creole ASR
%A Andrade, Mateus Neves
%A Ba, Mouhamadou Lamine
%A Diop, Idy
%A da Veiga, Arlindo Oliveira
%Y Matfunjwa, Muzi
%Y Setaka, Mmasibidi
%Y Mabuya, Rooweither
%Y van Zaanen, Menno
%S Proceedings of Resources for African Indigenous Languages (RAIL) 2026 @ LREC 2026
%D 2026
%8 May
%I ELRA Language Resources Association (ELRA)
%C Palma, Mallorca (Spain)
%F andrade-etal-2026-less
%X Automatic Speech Recognition (ASR) for low-resource languages remains challenging due to limited annotated data and high linguistic variability. In this work, we investigate parameter-efficient fine-tuning strategies for Cape Verdean Creole ASR using the Wav2Vec 2.0 XLSR model. We evaluate the impact of structured layer freezing on model performance, training stability, and computational efficiency. Experiments conducted on a newly curated Santiago-dialect dataset show that full fine-tuning achieves the best absolute performance (WER 0.212, CER 0.120). However, several freezing configurations achieve comparable recognition performance while substantially reducing the number of trainable parameters and exhibiting more stable convergence. These results highlight a trade-off between adaptability and efficiency, showing that selective freezing can serve as an effective regularization strategy in low-resource settings. This work provides practical insights into parameter-efficient adaptation for under-resourced Creole languages.
%R 10.63317/4b6ij7awds27
%U https://aclanthology.org/2026.rail-1.7/
%U https://doi.org/10.63317/4b6ij7awds27
%P 62-71
Markdown (Informal)
[Less can be More: Towards a Parameter-Efficient Fine-Tuning of Wav2Vec2 XLSR for Low-Resource Cape Verdean Creole ASR](https://aclanthology.org/2026.rail-1.7/) (Andrade et al., RAIL 2026)
ACL