@inproceedings{sankar-etal-2025-towards,
title = "Towards Building Large Scale Datasets and State-of-the-Art Automatic Speech Translation Systems for 14 {I}ndian Languages",
author = "Sankar, Ashwin and
Jain, Sparsh and
Narasimhan, Nikhil and
Choudhary, Devilal and
Suman, Dhairya and
Khan, Mohammed Safi Ur Rahman and
Kunchukuttan, Anoop and
Khapra, Mitesh M and
Dabre, Raj",
editor = "Che, Wanxiang and
Nabende, Joyce and
Shutova, Ekaterina and
Pilehvar, Mohammad Taher",
booktitle = "Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)",
month = jul,
year = "2025",
address = "Vienna, Austria",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2025.acl-long.1582/",
doi = "10.18653/v1/2025.acl-long.1582",
pages = "32945--32966",
ISBN = "979-8-89176-251-0",
abstract = "Speech translation for Indian languages remains a challenging task due to the scarcity of large-scale, publicly available datasets that capture the linguistic diversity and domain coverage essential for real-world applications. Existing datasets cover a fraction of Indian languages and lack the breadth needed to train robust models that generalize beyond curated benchmarks. To bridge this gap, we introduce BhasaAnuvaad, the largest speech translation dataset for Indian languages, spanning over 44 thousand hours of audio and 17 million aligned text segments across 14 Indian languages and English. Our dataset is built through a threefold methodology: (a) aggregating high-quality existing sources, (b) large-scale web crawling to ensure linguistic and domain diversity, and (c) creating synthetic data to model real-world speech disfluencies. Leveraging BhasaAnuvaad, we train IndicSeamless, a state-of-the-art speech translation model for Indian languages that performs better than existing models. Our experiments demonstrate improvements in the translation quality, setting a new standard for Indian language speech translation. We will release all the code, data and model weights in the open-source, with permissive licenses to promote accessibility and collaboration."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="sankar-etal-2025-towards">
<titleInfo>
<title>Towards Building Large Scale Datasets and State-of-the-Art Automatic Speech Translation Systems for 14 Indian Languages</title>
</titleInfo>
<name type="personal">
<namePart type="given">Ashwin</namePart>
<namePart type="family">Sankar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sparsh</namePart>
<namePart type="family">Jain</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Nikhil</namePart>
<namePart type="family">Narasimhan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Devilal</namePart>
<namePart type="family">Choudhary</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Dhairya</namePart>
<namePart type="family">Suman</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mohammed</namePart>
<namePart type="given">Safi</namePart>
<namePart type="given">Ur</namePart>
<namePart type="given">Rahman</namePart>
<namePart type="family">Khan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Anoop</namePart>
<namePart type="family">Kunchukuttan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mitesh</namePart>
<namePart type="given">M</namePart>
<namePart type="family">Khapra</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Raj</namePart>
<namePart type="family">Dabre</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2025-07</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)</title>
</titleInfo>
<name type="personal">
<namePart type="given">Wanxiang</namePart>
<namePart type="family">Che</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Joyce</namePart>
<namePart type="family">Nabende</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Ekaterina</namePart>
<namePart type="family">Shutova</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Mohammad</namePart>
<namePart type="given">Taher</namePart>
<namePart type="family">Pilehvar</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Vienna, Austria</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-251-0</identifier>
</relatedItem>
<abstract>Speech translation for Indian languages remains a challenging task due to the scarcity of large-scale, publicly available datasets that capture the linguistic diversity and domain coverage essential for real-world applications. Existing datasets cover a fraction of Indian languages and lack the breadth needed to train robust models that generalize beyond curated benchmarks. To bridge this gap, we introduce BhasaAnuvaad, the largest speech translation dataset for Indian languages, spanning over 44 thousand hours of audio and 17 million aligned text segments across 14 Indian languages and English. Our dataset is built through a threefold methodology: (a) aggregating high-quality existing sources, (b) large-scale web crawling to ensure linguistic and domain diversity, and (c) creating synthetic data to model real-world speech disfluencies. Leveraging BhasaAnuvaad, we train IndicSeamless, a state-of-the-art speech translation model for Indian languages that performs better than existing models. Our experiments demonstrate improvements in the translation quality, setting a new standard for Indian language speech translation. We will release all the code, data and model weights in the open-source, with permissive licenses to promote accessibility and collaboration.</abstract>
<identifier type="citekey">sankar-etal-2025-towards</identifier>
<identifier type="doi">10.18653/v1/2025.acl-long.1582</identifier>
<location>
<url>https://aclanthology.org/2025.acl-long.1582/</url>
</location>
<part>
<date>2025-07</date>
<extent unit="page">
<start>32945</start>
<end>32966</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Towards Building Large Scale Datasets and State-of-the-Art Automatic Speech Translation Systems for 14 Indian Languages
%A Sankar, Ashwin
%A Jain, Sparsh
%A Narasimhan, Nikhil
%A Choudhary, Devilal
%A Suman, Dhairya
%A Khan, Mohammed Safi Ur Rahman
%A Kunchukuttan, Anoop
%A Khapra, Mitesh M.
%A Dabre, Raj
%Y Che, Wanxiang
%Y Nabende, Joyce
%Y Shutova, Ekaterina
%Y Pilehvar, Mohammad Taher
%S Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)
%D 2025
%8 July
%I Association for Computational Linguistics
%C Vienna, Austria
%@ 979-8-89176-251-0
%F sankar-etal-2025-towards
%X Speech translation for Indian languages remains a challenging task due to the scarcity of large-scale, publicly available datasets that capture the linguistic diversity and domain coverage essential for real-world applications. Existing datasets cover a fraction of Indian languages and lack the breadth needed to train robust models that generalize beyond curated benchmarks. To bridge this gap, we introduce BhasaAnuvaad, the largest speech translation dataset for Indian languages, spanning over 44 thousand hours of audio and 17 million aligned text segments across 14 Indian languages and English. Our dataset is built through a threefold methodology: (a) aggregating high-quality existing sources, (b) large-scale web crawling to ensure linguistic and domain diversity, and (c) creating synthetic data to model real-world speech disfluencies. Leveraging BhasaAnuvaad, we train IndicSeamless, a state-of-the-art speech translation model for Indian languages that performs better than existing models. Our experiments demonstrate improvements in the translation quality, setting a new standard for Indian language speech translation. We will release all the code, data and model weights in the open-source, with permissive licenses to promote accessibility and collaboration.
%R 10.18653/v1/2025.acl-long.1582
%U https://aclanthology.org/2025.acl-long.1582/
%U https://doi.org/10.18653/v1/2025.acl-long.1582
%P 32945-32966
Markdown (Informal)
[Towards Building Large Scale Datasets and State-of-the-Art Automatic Speech Translation Systems for 14 Indian Languages](https://aclanthology.org/2025.acl-long.1582/) (Sankar et al., ACL 2025)
ACL
- Ashwin Sankar, Sparsh Jain, Nikhil Narasimhan, Devilal Choudhary, Dhairya Suman, Mohammed Safi Ur Rahman Khan, Anoop Kunchukuttan, Mitesh M Khapra, and Raj Dabre. 2025. Towards Building Large Scale Datasets and State-of-the-Art Automatic Speech Translation Systems for 14 Indian Languages. In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 32945–32966, Vienna, Austria. Association for Computational Linguistics.