Jan is a principal research scientist working on topics at the cross-section of rendering and machine learning. He joined NVIDIA Research in 2018 after five years at Disney Resarch where he led the Rendering Group and worked on various techniques for efficient rendering and physically based simulation of light transport. He initiated and led several projects employing deep learning for image denoising, radiance estimation, and importance sampling, and co-developed techniques for rendering scenes with participating media. Some of his ideas have been integrated into Renderman and Hyperion renderers and are used in production at Pixar, Walt Disney Animation Studios, and ILM. Prior to Disney Research, he worked at the Walt Disney Animation Studios contributing to the Hyperion renderer. Jan holds a Ph.D. degree (2014) from the Karlsruhe Institute of Technology and B.Sc. (2007) and M.Sc. (2009) degrees from the Czech Technical University in Prague.

If not occupied with work, he’s busy with training four remarkably hard to train, real neural networks.