Working towards unravelling the limitation of pre-trained large models, specifically the sparse knowledge in them under the guidance of Prof. Yee Soon Ong.
Multi Media Laboratory @ Nanyang Technological University
Part-time Research Assistant Aug ’23 - Present
Working on preparing a comprehensive survey related to video-text pre-trained large models under the guidance of Prof. Ziwei Liu and Mr. Jingkang Yang.
University of Illinois Urbana-Campaign
Research Internship - CV, NLP, USA Sep ’21 - Present
Worked in collaboration with Dr. Haohan Wang and Mr. Haoyang Liu to write a survey titled </em>Towards Trustworthy and Aligned Machine Learning: A Data-centric Survey with Causality Perspectives</em>.
After an extensive read of literature on causality and pre-trained language models, developed novel master equations, to sum up hundreds of works from a causal perspective and writing about them.
Helped in developing the pre-trained model section highlighting their strength and weaknesses from a trustworthy and causal view.
MIT Driverless
Computer Vision Engineer - CV, USA Aug ’21 to Nov ’21
Closely worked with the team to implement the modules, based on cutting edge research, and attended Indy Autonomous Challenge.
A* Star
Research Internship - CV, NLP, Singapore Mar ’21 - Aug ’21
Worked with Dr. Shitala Prasad on reasearch paper titled “Harnessing Memory-Augmented Network for Caption Generation” by introducing an external memory to eradicate long-tailed bias and memory assistance for long-sequence generation.
Developed the idea and built the system on a Caption Generation system from scratch.
Indian Institute of Technology (IIT), Indore
Research Internship - ML, CV, India Oct ’19 - Mar’ 21
Proposed novel algorithm for detection module more robust by eradicating the confounding nature of the background in images, allowing the model to be deployed in the real world with an unwavering performance.
For increasing the performance, incorporated and implemented state-of-the-art Stand alone Self Attention research paper from scratch.
Helped to formulate pipeline and build dataset, by scraping 15k reviews from Tripadvisor.com and annotated them.
Jaypee University of Information Technology, Solan
Research Internship - ML, India Sep ’19 - Jun ’20
Worked with Prof. Mrityunjay Singh on project sponsored by Govt. of India to prepare the research work titled: “Recommendation Engine for Research Papers” and is under review.
Implemented a text embedding based sorting algorithm that is able to give recommendation to a user based on his/her expertise in the research domain.
Awards and Achievement
Selected as Mentor in UNESCO-India-Africa.
Held between 20+ countries of Africa and India; aimed to develop Voice assisted system for farmers.
Team Leader in ASEAN-India Hackathon
Held between 10+ south-east Asian countries; focused on solving the problem of Marine Species Detection under Blue Economy
Winner of Smart India Hackathon−world’s biggest hackathon
With an initial pool size of 0.2 million students. Worked with Bureau of Police Research & Development problem statement to identify the Face of Criminals.
Twitter
LinkedIn
Github
Google Scholar