The initiatives, aimed at making the company’s tools more comprehensive and inclusive will rely heavily on the use of artificial intelligence (AI) and machine learning (ML), the company’s executives said.
In one of its most ambitious projects in India, the company has tied up with the Indian Institute of Science for Project Vaani under which it will collect and analyse voice samples from 80 districts initially to train the voice search. The project will eventually expand to cover all 773 districts of India.
“Our mission is to democratise access to information to every person in their native language. Voice has a very key role to play in India,” Manish Gupta, the director of Google Research in India told ET.
The idea of the project, Gupta said, is to capture the variety and richness of Indian language diaspora and ensure that any language in India that is spoken by at least one lakh people is included in the research project. Google will also make the raw data of the project open source so that it is available to the industry and academia for any other research or use, Gupta said.
Further, Google plans to augment voice and video search in India, where the adoption of these two is twice the world average, Elizabeth Reid, the vice-president of Engineering at Google told ET.
Discover the stories of your interest
The idea, Reid said, was to bring the capabilities of voice and video search in a user’s own language in such a way that it would have the “ease of talking to someone at home and asking them a question”.
“People in India consume language with a lot of complexity. All of us at home speak at home using Hindi words and some regional languages. We want people to be able to embrace and interact with technology just like they talk at home. I should not learn how to interact with technology. Rather, technology should learn how I like to interact. Because, how I like to interact is different than how you like to interact,” Puneesh Kumar the general manager of search at Google India told ET.
Google will also expand the use of image search in a way that will allow users to point their smartphone camera at objects and get more comprehensive information about them. Not just this, users will also be able to search for and directly jump to a particular section of a YouTube video with the video search option. The beta version of this feature is presently being tested, Reid said.
Google also plans to use its AI capabilities extensively in remote sensing and analysing of farmlands to provide details such as the nature of the soil, the crop that should be cultivated, the crop that is currently being cultivated on that strip of land, as well as other details such as whether a water body present near the farmland is natural or man-made.
“We are taking a very partnerships-based approach. We work on the basic foundational layers and make them available to startups, policymakers and so on. Many of these startups, agri-tech companies build solutions on top of that which are being given to farmers, banks as well as policymakers,” Gupta said.
Further, Google has also expanded the use of AI to launch the pilot of Project Relate, an app trained to ‘listen, repeat and assist’ people and users with non-standard speech. The project, which is being started with English testers in India, will be expanded to Hindi language users in early 2023, the company said.
The company’s video platform YouTube is piloting an AI-enabled solution Aloud, to allow some healthcare providers to simplify the process of dubbing and creating content that supports uploading a single video with multiple audio tracks.
To augment its capabilities in understanding healthcare, Google on Monday also unveiled an AI and ML model that can interpret handwritten medical prescriptions. The internet giant said this technology will let users understand handwritten prescriptions and help pharmacists digitise handwritten medical records.
The function is currently being developed, and no choice will be made fully based on the judgement of AI and ML technology, Google said.
In keeping with the theme of learning, YouTube said it was looking to provide more monetization opportunities to content creators and ease of access to a structured programme with its Courses feature which will make it compete with the likes of Byju’s and Unacademy.
Recognising the need for safe and secure access to documents on a user’s phone, Google announced the integration of Digilocker with the Files by Google app on Android. This will allow users to easily access their key digital documents privately and securely and will help to ensure that important information is secure and within easy reach.
It also launched a new fraud detection model for digital payments on Google Pay.
The Google for India Digitization Fund will provide support for early-stage and women-led start-ups in the country to further drive innovation and entrepreneurship in the space.
Google will give a $1 million grant to the Indian Institute of Technology in Madras to set up India’s first Centre for Responsible AI, which will focus on research, education, and outreach in the field.
It also announced a $1 million grant via Google.org to support the deployment of AI models across crop disease monitoring, predicting yield outcomes, and bringing efficiencies to Kisan call centres with Wadhwani AI.
For all the latest Technology News Click Here
For the latest news and updates, follow us on Google News.
