Tomáš Souček is a postdoctoral researcher at Meta FAIR with a decade of experience building weakly supervised 3D computer vision systems and occasional CUDA optimizations to accelerate Python ML/CV stacks. He leads research on in-model and out-of-model watermarking for autoregressive and diffusion models, work that is deployed in production across Instagram and Meta’s AI app to authenticate millions of images. Author and lead author on multiple CVPR, TPAMI and ACM MM papers, he also contributes to ICLR and ICML workshop publications and brings a PhD focused on weak supervision from Czech Technical University. Based in Paris, he combines deep academic rigor with production-minded engineering, uniquely bridging foundational vision research and scalable deployment.
10 years of coding experience
Doctor of Philosophy - PhD, Weakly supervised learning for visual recognition, Doctor of Philosophy - PhD, Weakly supervised learning for visual recognition at Czech Technical University in Prague
Master's degree, Artificial Intelligence, summa cum laude, Master's degree, Artificial Intelligence, summa cum laude at Charles University in Prague
Contributions:7 commits, 6 pushes, 1 branch in 1 year
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.