Hi! I am a PhD student in the Department of Computer Science at University of Toronto, where I am advised by Igor Gilitschenski. I am currently working on neural rendering for 3D scenes. Previously, I was a Pre-Doctoral researcher at the Earth Observation Science group at Google Research, India where I worked with Varun Gulshan . My research there focused on self-supervised learning with large-scale multimodal satellite data for remote sensing applications.

I graduated from Indian Institute of Technology Madras with a B.Tech (Hons.) and a M.Tech in Electrical Engineering, and minors in Mathematics and Computing. I completed my thesis with Prof. Harish Guruprasad on generalization in deep neural networks. I also worked in the Computational Neuroscience Lab at IIT Madras. My research interests lie in applying machine learning for computer vision and multimodal content understanding.

Publications

GaussianCut: Interactive segmentation via graph cut for 3D Gaussian Splatting
Umangi Jain, Ashkan Mirzaei, Igor Gilitschenski
NeurIPS, 2024

Satellite Sunroof: High-res Digital Surface Models and Roof Segmentation for Global Solar Mapping
Vishal Batchu, Alex Wilson, Betty Peng, Carl Elkin, Umangi Jain, Christopher Van Arsdale, Ross Goroshin, Varun Gulshan
NeurIPS Workshop on Tackling Climate Change with Machine Learning, 2024

Detangling the role of climate in vegetation productivity with an explainable convolutional neural networks
Ricardo Barros Lourenço, Michael J Smith, Sylvia Smullin, Umangi Jain, Alemu Gonsamo, Arthur Ouaknine
Tech Report (Climate Change and AI Summer School'23)

Multimodal contrastive learning for remote sensing tasks
Umangi Jain, Alex Wilson, Varun Gulshan NeurIPS Workshop on Self-Supervised Learning - Theory and Practice (SSL@NeurIPS)

Convolutional elman jordan neural network for reconstruction and classification using attention window
Sweta Kumari, S Aravindakshan, Umangi Jain, V Srinivasa Chakravarthy Innovations in Computational Intelligence and Computer Vision (ICICV'20 )