Scholar
Chaitanya Devaguptapu
Google Scholar ID: m06B1c4AAAAJ
IIT Hyderabad
Citations & Impact
Citations
314
H-index
5
i10-index
5
Publications
14
Co-authors
0
Publications
Resume (English only)
Academic Achievements
- Three papers accepted at EMNLP-2025; one paper on 'Finding Needles in Images' accepted at ACL-2025 and received a Highlights award; one paper on Hybrid Graphs for Table-Text QA accepted at NAACL-2025; one paper on a training-free approach for 3D Scene Editing accepted at WACV-2025; one paper on improving annotation efficiency accepted at ICML workshops; one paper on adversarial robustness from a neural architecture search perspective accepted at the Adversarial Robustness in the Real World workshop at ICCV-21; served as a reviewer for multiple top conferences and journals such as ICLR, AAAI, WACV, NeurIPS, and IEEE Pattern Recognition Journal.
Research Experience
- Currently a Researcher at Fujitsu Research India, leading a small team on GenAI research. Served as a Teaching Assistant at IIT-Hyderabad and as a mentor and coach with platforms like Udacity and UpGrad.
Education
- Master’s by Research (M.Tech-RA) in Computer Science from IIT-Hyderabad, advised by Prof. Vineeth N Balasubramanian. Worked as a visiting researcher at the University of Toronto and Vector Institute, advised by Prof. Animesh Garg.
Background
- AI Researcher focusing on Generative AI and related applications. Interested in problems with clear practical applications or purely exploratory endeavors that spark curiosity. Explored data-efficient learning techniques, including active learning, transfer learning, and multi-modal learning during his Master's. Currently, his work spans the broader spectrum of Generative AI, combining research insights with engineering to deliver useful solutions. Maintained a strong connection to education and teaching throughout his academic career.
Miscellany
- Personal interests include education and teaching, having served as a Teaching Assistant at IIT-Hyderabad and as a mentor and coach with online education platforms like Udacity and UpGrad.
Co-authors