Project Vaani : How Google aims to comprehend all dialects and languages spoken in India

Project Vaani: How Google aims to comprehend all dialects and languages spoken in India

Project Vaani, a collaborative project of IISc, Google and ARTPark to collect speech samples from all over India in order to build a language model powered by AI that can comprehend the various Indian languages and dialects.

What is Project Vaani?

• As part of Project Vaani, voice sets from over 1 million people in 773 districts over the course of three years will be collected in order to map the many languages spoken throughout India.
• This project is expected to cost between $30 and $40 million USD.
• It is a component of the Bhasha AI project from the IISc and Artpark in Bengaluru, which also consists of RESPIN (Recognizing Speech in Indian languages) and SYSPIN (Synthesizing Speech in Indian languages).
• In order to complete the project, IISc and Google will record approximately 1.5 lakh hours of speech, some of which will be transcribed into local characters.
• This project employs a district-anchored methodology, which entails selecting more than 1,000 individuals at random from each district to record local remarks.

Objectives of Project Vaani

• The development of technologies like automatic voice recognition, speech-to-speech translation, and natural language understanding is one of the key goals of this project.
• The ultimate objective is to provide a technological solution that can do rid of the linguistic barriers that are currently in place in technology and make it more accessible to a larger range of individuals.
• After this project is finished, attempts will be made to develop a language model powered by artificial intelligence that can comprehend the various Indian languages and dialects.
• The Vaani project’s new approach allows both speech and text translation. The Multilingual Representations for Indian Languages (MuRIL), which only enables text-based translation, might be improved by this. Over 100 Indian languages, spoken by more than 1 lakh people nationwide, would be used to train the new model.

Current Status of the Project Vaani

• Language information from roughly 69 districts in India has been gathered during the previous few months.
• More than 150 hours of data have already been gathered, evenly distributed by gender and age, from 841 distinct pin codes in more than 30 different languages.

MPPSC Free Study Material (English/Hindi)

🤩Follow Our Social Media Handles

YouTube 👉 https://bit.ly/36wAy17
Telegram👉 https://bit.ly/3sZTLzD
Facebook 👉 https://bit.ly/3sdKwN0

Daily Current Affairs Quiz for UPSC, MPSC, BPSC, and UPPSC: Click here