UrHiOdSynth: A Multilingual Synthetic Corpus for Speech-to-Speech Translation in Low-Resource Indic Languages

Abstract

Speech-to-Speech Translation (S2ST) focuses on generating spoken output in a target language directly from spoken input in a source language. Despite progress in S2ST modeling, low-resource Indic languages remain poorly supported, primarily because large-scale parallel speech corpora are unavailable. We present UrHiOdSynth, a three-language parallel S2ST synthetic dataset containing approximately 75 hours of speech across Urdu, Hindi, and Odia. The synthetic corpus consists of 10,735 aligned sentence triplets, with an average utterance length of 8.45 seconds. To our knowledge, UrHiOdSynth represents the largest multi-domain resource offering aligned speech and text for S2ST in this language context. Beyond speech-to-speech translation, the dataset supports tasks such as automatic speech recognition, speech-to-text translation, text-to-speech synthesis, and machine translation. This flexibility enables the fine-tuning of unified multilingual models, particularly for low-resource Indic languages. The dataset and code are publicly available at https://github.com/UrHiOdsynth/UrHiOdsynth. © 2026 Association for Computational Linguistics.