Akash Gupta

Senior Machine Learning Scientist

Vimaan AI

About Me

Welcome to my personal website!

My name is Akash Gupta, and I am currently working as Senior Machine Learning Scientist at Vimaan AI. Prior to joining Vimaan AI, I pursued my Ph.D. at University of California, Riverside under the supervision of Prof. Amit Roy-Chowdhury from 2018 to 2021 with thesis on Video Enhancement using Internal Learning and Blind Priors. I also worked on face recognition in art portraits as part of my Master's thesis from 2016 to 2018.

Before coming to UCR, I received my B. Tech in Electronics and Communications from Visvesvaraya National Institute of Technology, Nagpur, India in 2014.

Outside of research, I enjoy sketching, poetry, cooking, running, movies and travelling.

Download my CV.

Interests

Computer Vision
Video Synthesis and Enhancement
Video Super-Resolution

Education

PhD in Electrical Engineering, 2018-2021
University of California, Riverside
MS in Electrical Engineering, 2016-2018
University of California, Riverside
B. Tech. in Electronics and Communication Engineering, 2010-2014
Visvesvaraya National Institute of Technology, Nagpur, India

News

Paper on joint rolling shutter correction and super-resolution accepted in WACV 2023

Oct 21, 2022

Paper on multi-object adversarial attack accepted to NeurIPS 2022

Sep 15, 2022

Paper on self-supervised image enhancement accepted to MICCAI 2022

Jun 3, 2022

Successfully defended my thesis on "Video Enhancement using Internal Learning and Blind Priors"

Sep 17, 2021

Paper on adaptive video super-resolution accepted as Oral in ACM-MM 2021

Aug 12, 2021

See all events

Recent Publications

Joint Video Rolling Shutter Correction and Super-Resolution

We propose PatchNet that bi-directional motion field in the feature space to extract relevant information from neighboring patches using attention mechanism, and deformable fields using deformable convolutions to extract local pixel-level information for joint rolling shutter correction and super-resolution.

In IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), 2023
Akash Gupta, Sudhir Singh, Amit K. Roy-Chowdhury

Joint Video Rolling Shutter Correction and Super-Resolution

GAMA: Generative Adversarial Multi-Object Scene Attacks

GAMA demonstrates the utility of the CLIP model as an attacker’s tool to train formidable perturbation generators for multi-object scenes. Using the joint image-text features to train the generator, we show that GAMA can craft potent transferable perturbations in order to fool victim classifiers in various attack settings.

In Neural Information Processing Systems (NeurIPS), 2022
Abhishek Aich, Calvin-Khang Ta, Akash Gupta, Chengyu Song, Srikanth V Krishnamurthy, M Salman Asif, Amit K Roy-Chowdhury

GAMA: Generative Adversarial Multi-Object Scene Attacks

Poisson2Sparse: Self-Supervised Poisson Denoising From a Single Image

We propose a self-supervised learning methos that utilizes sparsity and dictionary learning-based approach for single-image denoising where the noise is approximated as a Poisson process, requiring no clean ground-truth data.

In The Medical Image Computing and Computer Assisted Intervention (MICCAI), 2022
Calvin-Khang Ta, Abhishek Aich, Akash Gupta, Amit K. Roy-Chowdhury

Poisson2Sparse: Self-Supervised Poisson Denoising From a Single Image

Ada-VSR: Adaptive Video Super-Resolution with Meta-Learning

We propose Ada-VSR framework that employs meta-learning to obtain adaptive model parameters, using a large-scale external dataset, that can adapt quickly to novel conditions of the given test video, thereby exploiting external and internal information of a video for super-resolution.

In ACM International Conference on Multimedia (ACM-MM), 2021 (Oral Presentation)
Akash Gupta, Padmaja Jonnalagedda, Bir Bhanu, Amit K. Roy-Chowdhury