Avijit Dasgupta

I am a M.S. + Ph.D. student in Computer Science at IIIT Hyderabad, India and a member of the CVIT lab, jointly advised by C. V. Jawahar, IIIT Hyderabad, and Karteek Alahari at the Thoth group, Inria. My research focuses on developing video understanding models with the ability to generalize effectively across previously unseen domains. One of our projects is featured on Meta's YouTube channel. Check it out here.

In 2019, I've spent three wonderful months as a visiting researcher at the Thoth group, Inria. My Ph.D. is funded by the Google India Ph.D. Fellowship, 2017 (1 out of 4 recipients). Thank you Google!

News

Education

Publications

Pattern Recognition 2025

Source-free video domain adaptation by learning from noisy labels

Avijit Dasgupta, C. V. Jawahar, Karteek Alahari

PaperCode

Source-free video adapation by a teacher-student framework

CVPR 2024

Oral Presentation

Ego-Exo4D: Understanding Skilled Human Activity from First- and Third-Person Perspectives

Kristen Grauman, Andrew Westbury, Lorenzo Torresani, Kris Kitani, Jitendra Malik, ..., Avijit Dasgupta* et al.

PaperCode

A diverse, large-scale multimodal multiview video dataset capturing skilled human activities

ICVGIP 2022

🏆 Best Paper Runner-up Award

Overcoming Label Noise for Source-free Unsupervised Video Domain Adaptation

Avijit Dasgupta, C. V. Jawahar, Karteek Alahari

PaperCode

A self-training based approach for source-free video domain adaptation

ICPR 2020

Context Aware Group Activity Recognition

Avijit Dasgupta, C. V. Jawahar, Karteek Alahari

Paper

Combines person-level appearance features with contextual cues (pose and scene context) to enrich representations.

ISBI 2017

A Fully Convolutional Neural Network based Structured Prediction Approach Towards the Retinal Vessel Segmentation

Avijit Dasgupta*, Sonam Singh*

Paper

A CNN-based approach for retinal blood vessel segmentation.

Experience