Aryan Garg

Currently, I'm a first year Computer Science PhD student at UW-Madison, advised by Professor Mohit Gupta.

For the summer of '24, I interned at the Computational Behaviour Lab directed by Prof. Antti Oulasvirta at Aalto University. I was extremely fortunate to have close support from Dr. Yue Jiang.

Last, I was a Bosch AI Research Fellow at the Computational Imaging Lab directed by Prof. Kaushik Mitra at the Indian Institute of Technology, Madras. In a past life, I finished my Bachelor's in Computer Science at IIT Mandi, where I was advised by Prof. Renu Rameshan. During my junior year, I had the privilege of being advised by Prof. Jean-François Lalonde and Dr. Yannick Hold-Geoffroy while spending a MITACS summer at Université Laval in the beautiful Quebec City.

Apart from research and academic services, I enjoy reading 📚, traveling ✈️, competing in endurance sports 🚴‍♂️, and trekking 🏔️. My previous trek (8th) was Sandakphu-Phalut in the Himalayas overlooking the majestic sleeping Buddha.

Email  /  CV  /  Scholar  /  Github  /  Kaggle

Research Interests

My main focus is on deep generative modeling and 3D scene reconstruction. Outside of that, I also enjoy going on side quests 🗡️ to push the boundaries of core machine learning.

News

• Aug-19, 2024 Joined UW-Madison as a PhD student
• Jun-3, 2024 Interning this summer with Prof. Antti Oulasvirta at Aalto, Finland.
• May-16, 2024 Won a grant for CILAB worth 30,000 USD from ISRO to develop a rocket launch 3D visualization module.
• Apr-15, 2024 Accepted to PhD programmes at UW Madison, Rice University and KAUST.
• Nov-1, 2023 Awarded the Post-Bacc. Bosch AI Fellowship (RBCDSAI, IIT Madras)
• Jun-12, 2024 Joined CILab directed by Prof. Kaushik Mitra at IIT Madras.
• Jun-1, 2023 Graduated from IIT Mandi with Bachelors in CS.
• June 2022 Won the MITACS Award. Interning this summer with Prof. Jean Francois Lalonde at ULaval, Canada.

High Dynamic Range 3D Gaussian Splatting At Night

Stereo Knowledge Distillation for dp-disparity estimation, dpMV dataset, and LFVR.

Unconditional face generation at the speed and size of DCGAN using a relatively tiny training dataset (~4000). Convergence and higher fidelity over StyleGAN are achieved using a spectrally normalized discriminator and a tiny generator within 20 minutes on a commercial, free cloud GPU.

Synthetic dataset generation and sinusoidal activations lead to a winning solution over the final displacement error.

SkyNet for user-controlled generation of skies incorporating the versatility of parametric models and realism from deep generative networks.

DeepSky: Learning to generate photorealistic skies per user-controlled positioning of solar and atmospheric components

Academic Services

Reviewer For:

• ICLR 2025
• CVPR-W: AI for Content Creation, AI4CC 2024
• ICLR 2024
• (ICVGIP) 14th Indian Conference on Computer Vision, Graphics and Image Processing, 2023

Teaching Assistant

• CS639.1 - Data Management for Data Science (Prof. Meenakshi Syamkumar) - Fall '24, UW-Madison
• EE5719 - Deep Learning for Imaging (Prof. Kaushik Mitra) - Fall '23, IIT Madras (NPTEL)
• CS208 - Mathematical Foundations of Computer Science (Prof. Samar Agnihotri) - Fall '21, IIT Mandi
• EE522 - Matrix Theory for Engineers (Prof. Renu Rameshan) - Fall '21, IIT Mandi

More Socials