Summary
Koutilya Pnvr is an Applied Scientist at Adobe specializing in large-scale video foundation models for text-to-video, controllable video editing, and video super-resolution, leveraging a Ph.D. in Computer Vision from the University of Maryland. Over nine years he has bridged deep generative research—GANs and latent diffusion models—with production-focused engineering, driving advances in text-based image segmentation, domain adaptation for geometry estimation, and personalization of generative models. His work at Amazon and NEON demonstrated practical gains from repurposing LDM internal features to boost segmentation accuracy, and he now applies those insights to scalable video synthesis at Adobe Firefly. Comfortable across Python, C++ and CUDA stacks, he pairs rigorous academic methods with systems thinking to move prototypes into production. Based in San Jose, he actively seeks industry collaborations to push GenAI-forward video applications and often brings cross-disciplinary ideas from remote sensing and audio-visual projects into his research.
9 years of coding experience
2 years of employment as a software developer
Indian Institute of Technology Delhi (IIT Delhi)
Nanodegree, Deep Learning, Nanodegree, Deep Learning at Udacity
Doctor of Philosophy - PhD, Computer Vision, Doctor of Philosophy - PhD, Computer Vision at University of Maryland
English, Hindi, Telugu