Welcome
I'm an associate professor in Computer Science & Engineering
at Indian Institute of Technology, Kanpur.
Before this (reverse chronologically), I was a research faculty in
Electrical and Computer Engineering,
at Duke University,
a post-doctoral fellow in Statistics
and Computer Science
at UT Austin,
and a Ph.D. student in School of Computing
at University of Utah. I work in the
area of machine learning and Bayesian statistics. My research is primarily on probabilistic/generative
modeling and understanding of massive and complex data.
My students
and I focus on designing probabilistic generative models
and efficient optimization/inference algorithms that can learn compact and interpretable latent structures
and feature representations from such data, and leverage these to better understand/summarize the data,
generate data with controllable properties, and make better predictions/decisions, especially in the face of
data/model uncertainty. We also work on designing
machine learning models and algorithms that can work well in resource-constrained settings (e.g., scarce training data
or under limited compute/storage).
Some Recent Courses (more)
- CS771A: Introduction to Machine Learning (2024), (2023), (2020), (2018)
- CS772A: Probabilistic Machine Learning (2024) , (2023)
- CS698X: Topics in Probabilistic Modeling and Inference (2021), (2019)
- ESC101A: Fundamentals of Computing (2019)
Some Recent Papers (full list and Scholar page)
VERSE: Virtual-Gradient Aware Streaming Lifelong Learning with Anytime Inference
With:Soumya Banerjee, Vinay Verma, Avideep Mukherjee, Deepak Gupta, and Vinay Namboodiri
ICRA 2024RISSOLE: Parameter-efficient Diffusion Models via Block-wise Generation and Retrieval-Guidance
With:Avideep Mukherjee, Soumya Banerjee, and Vinay Namboodiri
BMVC 2024A Probabilistic Framework for Lifelong Test-Time Adaptation
With:Dhanajit Brahma
CVPR 2023SITA: Single Image Test-time Adaptation
With:Ansh Khurana, Sujoy Paul, Gaurav Aggarwal, and Soma Biswas
CVPR 2023 workshop on Computer Vision in the WildFederated Learning with Uncertainty via Distilled Predictive Distributions
With:Shrey Bhatt and Aishwarya Gupta
ACML 2023Gradient Perturbation-based Efficient Deep Ensembles
With:Amit Chandak and Purushottam Kar
IKDD-CODS 2023DiffuseVAE: Efficient, Controllable and High-fidelity Generation from Low-dimensional Latents
With:Kushagra Pandey, Avideep Mukherjee, and Abhishek Kumar
TMLR 2022Novel Class Discovery without Forgetting
With:K.J. Joseph, Sujoy Paul, Gaurav Aggarwal, Soma Biswas, Kai Han, Vineeth Balasubramanian
ECCV 2022Rectification-based Knowledge Retention for Task Incremental Learning
With:Pratik Mazumder, Pravendra, Singh, and Vinay P. Namboodiri
IEEE TPAMI 2022CAM-GAN: Continual Adaptation Modules for Generative Adversarial Networks
With:Sakshi Varshney, Vinay Verma, Srijith P.K., and Lawrence Carin
NeurIPS 2021Bayesian Structural Adaptation for Continual Learning
With: Abhishek Kumar and Sunabha Chatterjee
ICML 2021
Recent Service (External)
- Area Chair: ICML (2024, 2023, 2020), NeurIPS (2024, 2023, 2022), ICLR (2024, 2023, 2022), AAAI (2020, 2019), IJCAI (2020), ACML (2022)
- Senior Area Chair: AAAI (2024, 2023, 2022, 2021)
- Editorial boards: IEEE-TPAMI, ACM-ToPML
- Program Co-chair: Indian Workshop on Machine Learning (2018, 2016)
- Tutorial Co-chair: IKDD CODS-COMAD (2019)
Recent news (more)
- (October 2021) Our work on fine-grained emotion prediction, led by BT-MT student Gargi Singh, is the recipient of the best student paper award at ACII 2021. Congratulations, Gargi!
- (July 2021) Our work on semi-supervised continual learning, led by Ph.D. student Dhanajit Brahma, is the recipient of the best student paper award at CSSL@IJCAI 2021. Congratulations, Dhanajit!
- (June 2020) My Ph.D. student Dhanajit Brahma is one of the winners of Qualcomm Innovation Fellowship 2020. Congratulations, Dhanajit!
Contact
- Email: firstname AT cse DOT iitk DOT ac DOT in
- Office: RM-502, CSE Department, IIT Kanpur
- Phone: +91-512-259-6894