Language Data & AI
ARTPARK and IISc, along with partners, are working to create language data and AI that understands all Indians, so that Digital India is inclusive.
Imagine more and more people around you start speaking a new language you do not understand, and increasingly, they find it difficult to understand you too. How would you feel? Do you have to abandon the language you have been speaking all your life to be a part of the community again?
As essential services become digitized, there is a danger of hundreds of millions who are neither English speakers nor “digital natives” not being included.
BhashaSetu is an AI-powered API designed to bridge the language gap for applications that only support English. Without appropriate language AI, users are limited to interacting with the application solely in English. With BhashaSetu, the possibilities expand!
At ARTPARK/IISc, our belief is that language is a barrier that billions have to transcend to get access to the power and benefits unleashed by the internet and digital transformation. And language is, by default, diverse in nature. It changes with geography and over time. And our core belief is that helping billions benefit from the power of technology is best done via an ecosystem of partners. At the GPAI Event, ARTPARK/IISC is keen to demo the work we and our partners have done so far.
Datasets for an inclusive India
-
VAANI
Spoken language changes continuously with geography. It does not change abruptly at state or district boundaries. We are building pan-India mechanisms through which we are bring together speech and text data representing this diversity.
-
SYSPIN
Synthesizing Speech in Indian languages (SYSPIN) is an initiative to create a text-to-speech (TTS) synthesizer in nine Indian languages: Hindi, Bengali, Marathi, Telugu, Bhojpuri, Kannada, Magadhi, Chhattisgarhi and Maithili.
-
RESPIN
Speech recognition in agriculture and finance for the poor is an initiative to create resources and make them available as a digital public good in the open-source domain to spur research and innovation in speech recognition in nine different Indian: Hindi, Bengali, Marathi, Telugu, Bhojpuri, Kannada, Magadhi, Chhattisgarhi, and Maithili.
Applications for equitable use of AI
-
ARTPARK is a double winner of the BMGF AI Grand Challenges for Catalyzing Equitable AI Use.
-
LLM Copilot for Front Line Health Workers
A woman dies in childbirth every twenty minutes in India and for every woman who dies, 20 more suffer lifelong ailments. Over 26,000 women die from pregnancy/childbirth-related complications every year, out of ~30M pregnancies.
As a step to help train the front line health workers, ARMMAN, one of India’s largest NGOs for maternal and child healthcare,and ARTPARK are building an LLM-powered copilot into a learning & support app for an High Risk Pregnancy Management program that currently covers 10,000+ health workers. Better learning support and increased hand-holding will help health workers detect and manage high-risk pregnancy conditions early and effectively, leading to an overall reduction in maternal and neonatal mortality and morbidity.
This initiative is a winner of the Global AI Grand Challenges for Equiable AI by the Bill and Melinda Gates Foundation.
Notably, this project was featured amongst the 5 coolest innovations from across the globe by Mr. Bill Gates himself.
-
BelonggAI: Using LLMs to embed equity In SDG Research, Program Design, & Funding
Billions of people in LMICs with marginalized identities (gender, caste, disability, sexual orientation, ethnicity, religion) face suboptimal outcomes due to bias and poorly designed programs/policies that ignore their unique needs. These intersectional considerations are often overlooked in SDG activities, leaving gaps in addressing their needs.
BelonggAI and ARTPARK are developing an LLM-based tool to help development practitioners, funders, and researchers uncover and address these exclusions and make their work more inclusive of marginalized groups.
BelonggAI is like "Grammarly for equity & inclusion in SDG research and program design work".
This initiative is a winner of the Global AI Grand Challenges for Equiable AI by the Bill and Melinda Gates Foundation.