Amartya Dutta
PhD CS, Virginia Tech.

I am a PhD student in Computer Science at Virginia Tech, where I am co-advised by Dr. Anuj Karpatne and Dr. T.M. Murali. I am currently working on building foundation models at the COMPASS Centre. My research focuses on Language Models and multimodal AI. During my Master’s, I was a part of the KGML Lab and was advised by Dr. Anuj Karpatne. My thesis explored Vision-Language Models (VLMs) for predicting structured Scene Graph relationships without additional fine-tuning.
My academic journey began at the Indian Institute of Information Technology Guwahati (IIIT Guwahati), where I earned my Bachelor’s degree in Computer Science and Engineering under the guidance of Dr. Ferdous Ahmed Barbhuiya.
Outside of academia, I enjoy online gaming, soccer, and music. I am deeply passionate about singing and percussion, and I have been involved with bands such as Kala@VT and Ataasi The Band throughout my academic career.
Feel free to reach out if you’re interested in discussing AI research topics, collaborating on projects, or even jamming to some music!
News
Jun 12, 2025 | Open World Scene Graph Generation using Vision Language Models is out on arxiv. |
---|---|
Jun 09, 2025 | Two papers accepted accepted in ICML 2025 Workshop: Toward Scientific Foundation Models for Aquatic Ecosystems & Open World Scene Graph Generation using Vision Language Models. |
May 27, 2025 | Our paper Open World Scene Graph Generation using Vision Language Models is accepted in CVPR 2025 Workshop. |
May 10, 2025 | Two posters accepted in CVPR 2025 Workshop: Physics-guided Diffusion Neural Operators for Solving Forward and Inverse PDEs & Scientific Equation Discovery using Modular Symbolic Regression via Vision-Language Guidance. |
Mar 02, 2025 | ![]() |
Feb 27, 2025 | Successfully defended my Master’s Thesis titled Zero-Shot Scene Graph Relationship Prediction using VLMs at Virginia Tech! |
Jan 20, 2024 | ![]() |
Aug 20, 2023 | ![]() |
Jan 10, 2023 | ![]() |
Aug 22, 2022 | ![]() |