@inproceedings{oukid-etal-2026-puma,
title = "{PUMA}: Projected Universal Multilingual {ASR} for Low-Resource Settings. Application to Diverse {A}frican Languages",
author = "Oukid, Ilyes and
Faye, Bilal and
Azzag, Hanane and
Lebbah, Mustapha and
Boulahia, Said Yacine",
editor = "Liakata, Maria and
Moreira, Viviane P. and
Zhang, Jiajun and
Jurgens, David",
booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {ACL} 2026",
month = jul,
year = "2026",
address = "San Diego, California, United States",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.findings-acl.17/",
pages = "371--382",
ISBN = "979-8-89176-395-1",
abstract = "Multilingual ASR systems often fail to generalize to low-resource and linguistically diverse languages while remaining costly to scale. We introduce PUMA, a unified multilingual ASR model that improves low-resource performance with reduced model complexity. PUMA employs a Universal Language Projection (ULP) module that integrates a learnable language token with acoustic representations, enabling language-aware processing through shared parameters. Experiments on diverse African languages show consistent word error rate reductions over strong multilingual baselines, highlighting improved robustness and generalization. Our code is available at the following GitHub URL: https://github.com/ilyes-okd/PUMA"
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="oukid-etal-2026-puma">
<titleInfo>
<title>PUMA: Projected Universal Multilingual ASR for Low-Resource Settings. Application to Diverse African Languages</title>
</titleInfo>
<name type="personal">
<namePart type="given">Ilyes</namePart>
<namePart type="family">Oukid</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Bilal</namePart>
<namePart type="family">Faye</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hanane</namePart>
<namePart type="family">Azzag</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mustapha</namePart>
<namePart type="family">Lebbah</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Said</namePart>
<namePart type="given">Yacine</namePart>
<namePart type="family">Boulahia</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-07</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: ACL 2026</title>
</titleInfo>
<name type="personal">
<namePart type="given">Maria</namePart>
<namePart type="family">Liakata</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Viviane</namePart>
<namePart type="given">P</namePart>
<namePart type="family">Moreira</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jiajun</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">David</namePart>
<namePart type="family">Jurgens</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">San Diego, California, United States</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-395-1</identifier>
</relatedItem>
<abstract>Multilingual ASR systems often fail to generalize to low-resource and linguistically diverse languages while remaining costly to scale. We introduce PUMA, a unified multilingual ASR model that improves low-resource performance with reduced model complexity. PUMA employs a Universal Language Projection (ULP) module that integrates a learnable language token with acoustic representations, enabling language-aware processing through shared parameters. Experiments on diverse African languages show consistent word error rate reductions over strong multilingual baselines, highlighting improved robustness and generalization. Our code is available at the following GitHub URL: https://github.com/ilyes-okd/PUMA</abstract>
<identifier type="citekey">oukid-etal-2026-puma</identifier>
<location>
<url>https://aclanthology.org/2026.findings-acl.17/</url>
</location>
<part>
<date>2026-07</date>
<extent unit="page">
<start>371</start>
<end>382</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T PUMA: Projected Universal Multilingual ASR for Low-Resource Settings. Application to Diverse African Languages
%A Oukid, Ilyes
%A Faye, Bilal
%A Azzag, Hanane
%A Lebbah, Mustapha
%A Boulahia, Said Yacine
%Y Liakata, Maria
%Y Moreira, Viviane P.
%Y Zhang, Jiajun
%Y Jurgens, David
%S Findings of the Association for Computational Linguistics: ACL 2026
%D 2026
%8 July
%I Association for Computational Linguistics
%C San Diego, California, United States
%@ 979-8-89176-395-1
%F oukid-etal-2026-puma
%X Multilingual ASR systems often fail to generalize to low-resource and linguistically diverse languages while remaining costly to scale. We introduce PUMA, a unified multilingual ASR model that improves low-resource performance with reduced model complexity. PUMA employs a Universal Language Projection (ULP) module that integrates a learnable language token with acoustic representations, enabling language-aware processing through shared parameters. Experiments on diverse African languages show consistent word error rate reductions over strong multilingual baselines, highlighting improved robustness and generalization. Our code is available at the following GitHub URL: https://github.com/ilyes-okd/PUMA
%U https://aclanthology.org/2026.findings-acl.17/
%P 371-382
Markdown (Informal)
[PUMA: Projected Universal Multilingual ASR for Low-Resource Settings. Application to Diverse African Languages](https://aclanthology.org/2026.findings-acl.17/) (Oukid et al., Findings 2026)
ACL