NICT’s Cascaded and End-To-End Speech Translation Systems using Whisper and IndicTrans2 for the Indic Task

Raj Dabre, Haiyue Song


Abstract
This paper presents the NICT’s submission for the IWSLT 2024 Indic track, focusing on three speech-to-text (ST) translation directions: English to Hindi, Bengali, and Tamil. We aim to enhance translation quality in this low-resource scenario by integrating state-of-the-art pre-trained automated speech recognition (ASR) and text-to-text machine translation (MT) models. Our cascade system incorporates a Whisper model fine-tuned for ASR and an IndicTrans2 model fine-tuned for MT. Additionally, we propose an end-to-end system that combines a Whisper model for speech-to-text conversion with knowledge distilled from an IndicTrans2 MT model. We first fine-tune the IndicTrans2 model to generate pseudo data in Indic languages. This pseudo data, along with the original English speech data, is then used to fine-tune the Whisper model. Experimental results show that the cascaded system achieved a BLEU score of 51.0, outperforming the end-to-end model, which scored 19.1 BLEU. Moreover, the analysis indicates that applying knowledge distillation from the IndicTrans2 model to the end-to-end ST model improves the translation quality by about 0.7 BLEU.
Anthology ID:
2024.iwslt-1.3
Volume:
Proceedings of the 21st International Conference on Spoken Language Translation (IWSLT 2024)
Month:
August
Year:
2024
Address:
Bangkok, Thailand (in-person and online)
Editors:
Elizabeth Salesky, Marcello Federico, Marine Carpuat
Venue:
IWSLT
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
17–22
Language:
URL:
https://aclanthology.org/2024.iwslt-1.3
DOI:
10.18653/v1/2024.iwslt-1.3
Bibkey:
Cite (ACL):
Raj Dabre and Haiyue Song. 2024. NICT’s Cascaded and End-To-End Speech Translation Systems using Whisper and IndicTrans2 for the Indic Task. In Proceedings of the 21st International Conference on Spoken Language Translation (IWSLT 2024), pages 17–22, Bangkok, Thailand (in-person and online). Association for Computational Linguistics.
Cite (Informal):
NICT’s Cascaded and End-To-End Speech Translation Systems using Whisper and IndicTrans2 for the Indic Task (Dabre & Song, IWSLT 2024)
Copy Citation:
PDF:
https://aclanthology.org/2024.iwslt-1.3.pdf