About
I am a Sophomore at the Department of Computer Science and Engineering, IIT Kanpur. My interests broadly include Deep Learning and Cryptography.
I am associated with solvio.ai since mid-2020, where I am working on research involving Computer Vision and NLP. I work mostly on developing new techniques, often taking inspiration from existing research. A significant time of mine is spent implementing and experimenting models.
I also like to contribute to open source projects in Machine Learning, most of which I use. Last summer, I worked on multiple projects - Implementing popular Deep Learning models, developing the backend of a discussion forum using Django, studying new Cryptographic Primitives like Function Secret Sharing.
Currently, I am also spending time studying Reinforcement Learning, Quantum Computing, Probabilistic Machine Learning, and Cyber Security. I believe that (machine) learning should never stop, which motivates me to grow more every day. (pun intended!)
Interests
- Computer Vision
- Natural Language Processing
- Cryptography
- Graph Neural Networks
- Probabilistic Machine Learning
- Cyber Security
Education
B.Tech. CSE, 2023 (expected)
IIT Kanpur
Experience
AI Research and Development Intern
Solvio.ai
Secretary
Programming Club, IIT Kanpur
Projects
Model Zoo
Implementation of Deep Learning Models. We implemented 31 models in the Model-Zoo spanning categories like GANs, NLP, Classification, Multimodal Models, Audio Generation, Super Resolution and 3D Vision.
Campus Discuss Backend
A streamlined discussion forum for students. Developed using React and Django.
Private Computation using Cryptographic Primitives
Studied and explored applications of Cryptographic Primitives - Function Secret Sharing.
Probabilistic Machine Learning
Studied Probability Theory and Probabilistic Machine Learning, including Expectation Maximization, Variational Inference and Bayesian Regression.
gradeLess - A framework for AutoGrading
Developed a Web App using Angular on frontend, NodeJS and Python on backend to automatically grade objective type answers using Machine Learning algorithms. Part of ESC101A Course Project.