Academic StudyPublished: May 2023~512 citations recorded
IndicTrans2: Towards High-Quality and Accessible Machine Translation for all 22 Scheduled Indian Languages
Authors: Jay Gala, Pranjal A. Chitale, Raghavan AK, Mitesh M. Khapra
Institution / Lab: AI4Bharat, IIT Madras
Research Digest SponsorAdvertisement & Sponsorship
Paper Abstract & Theoretical Contribution
A monumental open-source machine translation architecture covering all 22 scheduled Indian languages across 44 combinations, drastically closing the low-resource benchmark gap for Indic digital sovereignty.
Key Experimental Findings & Benchmarks
Trained on Bharat Parallel Corpus (BPCC) comprising 230 million sentence pairs.
Outperformed Google Translate and Azure Cognitive Services across human evaluation metrics.
Serves as the multilingual linguistic foundation for IndiaAI and Bhashini public services.
Taxonomy & Field Classification:
#AI4Bharat#IIT Madras#Indic NLP#Machine Translation#Bhashini
Academic Network PlacementAdvertisement & Sponsorship