CVIT CVIT
CVIT CVIT
  • Home
  • People
    • Faculty
    • Staff
    • PhD Students
    • MS Students
    • Alumni
    • Post-doctoral
    • Honours Student
  • Research
    • Publications
    • Journals
    • Books
    • MS Thesis
    • PhD Thesis
    • Projects
    • Resources
  • Events
    • Talks and Visits
    • Major Events
    • Visitors
    • Summer Schools
  • Gallery
  • News & Updates
    • News
    • Blog
    • Newsletter
    • Past Announcements
  • Contact Us
  • Login
  1. You are here:  
  2. Home
  3. People
  4. PhD Students
  5. PhD Students
  6. Avijit Dasgupta

Avijit DasguptaAvijit Dasgupta

Areas of Interest:Computer Vision, Machine Learning (Deep Learning).
 
Email: This email address is being protected from spambots. You need JavaScript enabled to view it.
 
Address: CVIT, IIIT-H
 
Phone:
 
Personal Home Page: https://avijit9.github.io/

Publications

  • Distilling Distilling What and Why: Enhancing Driver Intention Prediction with MLLMs

    People Involved :  Sainithin Artham, Avijit Dasgupta, Shankar Gangisetty, and C. V. Jawahar

    Illustration of a driving scenario where the ADAS vehicle predicts a left lane change (what) to avoid slower traffic ahead (why). Existing DIP models lacking reasoning may miss such cues, while our framework jointly learns and distills both maneuver and explanation, improving decision quality.

     

  • Shayon Dasgupta, Avijit Dasgupta, and C V Jawahar -  Are We There Yet? Assessing the Capabilities of MLLMs in Assistive AI Applications, In Indian Conference on Vision Graphics and Image Processing (ICVGIP), 2025 [ PDF ]

  • Avijit Dasgupta,   C.V. Jawahar and Karteek Alahari -  Context Aware Group Activity Recognition  The  25th International Conference of Pattern Recognition  (ICPR) (ICPR 2021), Milano  [PDF]


Projects

Distilling Distilling What and Why: Enhancing Driver Intention Prediction with MLLMs

People Involved :  Sainithin Artham, Avijit Dasgupta, Shankar Gangisetty, and C. V. Jawahar

Illustration of a driving scenario where the ADAS vehicle predicts a left lane change (what) to avoid slower traffic ahead (why). Existing DIP models lacking reasoning may miss such cues, while our framework jointly learns and distills both maneuver and explanation, improving decision quality.

 

  • Avijit Dasgupta
  • Publications
  • Projects