Thomas Pock

Thomas Pock is a Professor of Computer Science at Graz University of Technology, holding the AIT Stiftungsprofessur for Mobile Computer Vision. He is affiliated with the Institute for Computer Graphics and Vision (ICG) within the Faculty of Computer Science and serves as a principal scientist at the Austrian Institute of Technology (AIT), Center for Vision, Automation & Control. He leads the Vision, Learning and Optimization (VLO) research group, which focuses on mathematical modeling and optimization in computer vision.

His research interests lie at the intersection of computer vision, image processing, and mathematical optimization. Specifically, he develops mathematical models for computer vision and efficient convex and non-smooth optimization algorithms, particularly for mobile scenarios. His recent work increasingly integrates variational methods with deep learning, especially in solving inverse problems in imaging such as medical reconstruction and deblurring.

The trends in his recent publications show a strong emphasis on deep learning for inverse problems, variational networks, and learned optimization. His group explores how to combine classical mathematical models with data-driven deep learning approaches to achieve stable, interpretable, and high-performance solutions in image reconstruction and processing.

His scientific achievements have been recognized with several prestigious awards:

• START Prize, Austrian Science Fund (FWF), 2013
• German Pattern Recognition Award, DAGM, 2013
• ERC Starting Grant, European Research Council, 2014

Thomas Pock actively mentors students and leads a research group of 10 PhD students and 2 postdocs. He has secured significant research grants, including the ERC Starting Grant, which supports his foundational work. He is also engaged in scientific communication, giving invited talks at international venues such as SIAM and co-organizing the IMAGINE One World seminar series to foster global collaboration in imaging and inverse problems.

He leads the Vision, Learning and Optimization (VLO) group at the Institute for Computer Graphics and Vision. The group develops mathematical models and efficient algorithms for computer vision and image processing, with a focus on mobile applications. The team includes multiple PhD students and postdoctoral researchers and has produced notable software and publications in top venues.