Scholar
Rohit Jena
Google Scholar ID: kZQQFE4AAAAJ
PhD in CS, University of Pennsylvania
Citations & Impact
Citations
354
H-index
11
i10-index
11
Publications
20
Co-authors
51
list available
Publications
Resume (English only)
Academic Achievements
- Selected Publications: WACV 2025 - 'Elucidating optimal reward-diversity tradeoffs in text-to-image diffusion models' (Work done at NVIDIA); NeurIPS 2024 - 'Deep Learning in Medical Image Registration: Magic or Mirage?'; CVPR - 'Beyond mAP: Towards better evaluation of instance segmentation'.
Research Experience
- Spent Summer 2024 at NVIDIA's NeMo team working on alignment of text-to-image diffusion models to improve the Pareto front of the alignment-diversity trade-off; Previously interned at Amazon Lab126, where he worked on mesh-NeRF hybrids for rigged 3D avatars from 360 degree videos.
Education
- Ph.D. student in CIS at University of Pennsylvania, advised by Prof. Pratik Chaudhari and Prof. James C. Gee; Master's from The Robotics Institute, Carnegie Mellon University, advised by Prof. Katia Sycara; Bachelor's in Computer Science and Engineering from Indian Institute of Technology, Bombay, with undergraduate thesis on 'Perfect Sampling and Uncertainty Estimation in Deep Networks' under the guidance of Prof. Suyash P. Awate.
Background
- Research interests include incorporating task-specific invariances for correspondence matching problems, discovering task-invariant representations through self-supervised learning, and the transferability of these representations. He believes in specialist models over generalist ones.
Miscellany
- The best way to reach him is via email.
