@ IIT Madras
The AI4Bharat Initiative
Mission statement of the Initiative
Bring parity in AI technology for Indian languages with English
Make fundamental contributions to state-of-the-art across language technologies - NLP, Speech, Sign language, OCR
Open source all datasets and AI models with permissible licenses
Build a coalition of partners to collect datasets and deploy models
Associate Professor, IIT Madras
PhD, IIT Bombay
Areas - NLP, Deep Learning
Mitesh M. Khapra
Pratyush Kumar
Our Team
Vivek Raghavan
Assistant Professor, IIT Madras
PhD, ETH Zürich
Areas - Deep Learning, Systems
AI Evangelist, EkStep Foundation
PhD, CMU
Areas - AI, Tech for social good
+ many hard-working students and volunteers
Project Sabdh Setu - Mission
Make educational content accessible in each of the 22 languages that are constitutionally recognised
A bridge to overcome the language divide
Aa
आ ஆ ଆ ...
Methodology
Follow the proven modern AI recipe
+
+
Huge amounts of data
Deep neural networks
Lot of compute power
+
Input & translate tools
Methodology
Follow the proven modern AI recipe
+
+
Huge amounts of data
Deep neural networks
Lot of compute power
+
Input & translate tools
What we have done so far
AI model for translating between English and 12 Indic languages
46M parallel sentences mined from the web (3X improvement)
Impact: The translation models (which are shown to be more accurate than commercial APIs) are being used to assist human translators in translating supreme court judgements with a significant increase in efficiency
Funding
What we have done so far
User types using English script
Automatically converted to Maithili script
Impact: Input tools become significantly more efficient for a long list of Indian languages. Impacts all content creation including writing storybooks for children
Deployed to write storybooks at
Funding
Outputs
Open source translation engine between English and all major Indian languages
Open source Android keyboards for 22 languages recognised in the constitution
Goal: Completely free and more accurate than commercial offerings
Outcomes
Direct impact on 20+ crore registered pupils
in schools in rural India
Especially important for digital access in times like COVID-19
Translate all NCERT textbooks from
Class I to XII into all major languages
Alignment with government
Alignment with government
Through EkStep, we are in close contact with the government on the NTM
SabdhSetu
Focus on higher education
Focus on primary education
Tools
AI models
Beyond Education
SabdhSetu
Supreme Court
of India
pilot ongoing
CDAC - evaluating for website translation
Ongoing field test to translate a fiction book
Deployed in
internal tool
Cost
