Failed to load scholar profile
Browse publications on Google Scholar (top-right) ↗
Resume (English only)
Academic Achievements
- Discovered that generative models can learn complex physical knowledge, such as intrinsic images—decomposing scenes into color, shape, and lighting—emerging within these models without explicit training [1,2,3]; identified fundamental limitations in their understanding of projective geometry, providing reliable detection signatures for generated images [4]; led community-building initiatives and organized workshops at conferences like CVPR and ECCV; LumiNet and Scribble Light accepted to CVPR 2025.
Research Experience
- Currently an Assistant Professor in the Department of Computer Science at Johns Hopkins University; formerly a Research Assistant Professor (RAP) at the Toyota Technological Institute at Chicago (TTIC); visiting scholar at UC Berkeley in Summer 2024, hosted by Alexei (Alyosha) Efros.
Education
- PhD from the University of Illinois Urbana-Champaign (UIUC), advised by David Forsyth; dual master’s degrees in Computer Science and Civil and Environmental Engineering at UIUC; bachelor’s degree in Civil Engineering from NITK Surathkal, India.
Background
- Research interests span computer vision, generative modeling, and physical reasoning. His goal is to develop, understand, and improve perception-driven and physics-aware generative models, thereby advancing the fields of computer vision, computer graphics, and computational photography.
Miscellany
- Designed and taught 'Past Meets Present: A Tale of Two Visions' at TTIC, exploring the intersections between classical computer vision and modern deep learning; wrote a new blog for junior grad students aiming to become faculty: 'So You Want to Be an Academic? What I Wish I Knew Early in Graduate School'.
Co-authors: 0 (list not available)