Amazon and Iit Bombay (IIT-B) Today, the initial award announced Hart Cape from the Amazon IIT-Bombay AI-ML initiative. The prizes recognize researchers whose work meets the initiative’s goals: to promote artificial intelligence and machine learning research in speech, language and multimodal-IA domains.
The Amazon-funded collaboration launched in March 2023 and the house in the IIT Bombay Department of Computer Science and Engineering aims to promote partnership between the faculty and leading scholars and promote another and sustainable pipeline of research talent.
In accordance with the initiative’s goals, the award winners are researching several domans, including Large Language Models (LLMS), Machine Learning, Federated Learning and Natural-Langage Processing (NLP).
“We are kindly wanting the unfolding of this relationship and the significant potential it is,” said Sachin Patwardhan, Dean of Research and Development at IIT Bombay. “The selection of nine project proposals under the Amazon IIT-B AI-ML initiative further highlights our common commitment to promote nowledge and innovation.”
The Research Award provides selected professors at IIT Bombay with up to an entire financing year to pursue independent research projects. The nine research projects selected will be run by the IIT Bombay faculty and researchers.
“We are excited about the progress we have made with our partnership with IIT-B,” said Snehal Meshram, senior manager for product management technology with Alexa. “Through the various project proposals, we have secured investments in key research areas that will mutually benefit Amazon and IIT-B. This is just the beginning of what we expect to be a long and fruitful collaboration.”
The winners of the awards are as follows:
Pushpak BhattacharyyaProfessor of Computer Science and Technology, and Preethi JyothiAssociate Professor of Computer Science and Technology
Bhattacharyya and Jyothi’s research are about translating speech-to-speech machine focusing on Indian languages. They intend to develop a multilingual and multimodal predicted model for Indian languages ​​and specific linguistic phenenels seen in India (such as Défluency and code mixing). The goal is a model that demoscers effectiveness in a speech-to-speech translation system, while the project will also notice curated training data specifically for the Indian use case.
Abir theyProfessor of computer science and engineering assistant, and Soumen ChakrabartiProfessor of Computer Science and Technology
From and Chakrabarti’s research focuses on multimodal representation, retrieval and transformation using graphructure objects. The proposal aims to develop methods for obtaining information on a large corpus of graphs, which will allow multimodal retrieval (text and images).
Opinion GhoshProfessor Assistant, Systems and Control Technology and Swaprava NathDeputy in computer science and technique
Ghosh and Nath’s proposals aim to utilize game-theoretical constructions to incentive contributions from Federated Learning users, where a large number of users submit locally calculated results merged into a centralized server.
Preethi JyothiAssociate Professor of Computer Science and Technology
Jyothi’s suggestions to build techniques that will make cross-language transfers more efficient and effective using both automatic-speaking (ASR) and NLP Autrained models.
Ankur A. KulkarniAssociate Professor of System Control and Technology
Kulkarni’s proposal aims to develop game-theoretical mechanisms for strategic classification, such as when input into a machine learning model has been targeted in test time.
Ajit RajwadeAssociate Professor of Computer Science and Technology
Rajwade’s research examines the closest neighbor search in large data sets via group testing. By combining more samples in a certain way, research will determine if fewer tests are needed.
Ganesh RamakrishnanProfessor, computer science and technique and Kshitij JadhavProfessor Assistant, Koita Center for Digital Health
Ramakrishnan and Jadhav’s research will use LLMs in the health field and explore both explainability and verifiableness. Their work will incorporate domain -specific training, increase in retrieval and adaptation to feedback from health experts.
Sunita SarawagiProfessor, Computer Science and Technology
Sarawigi’s research will investigate the integration of LLMs with structured databases. Her proposal aims to investigate a number of problem men, included the conversion of text to SQL, the continuous improvement of LLMs and related trends in structured data to realities in the real world, and to provide this capacity in custom LLMs developed over private data.
Vretra SinghProfessor, Department of Electrical Engineering
Singh’s research will use natural language to adapt reinforcement learning (RL) against human -like behavioral means. His proposal aims to develop technical to make RL more test -efficient using aids NLP tasks, such as providing text descriptions of the actions taken by RL agents.