Ulugbek Kamilov (WashU): Schmidt AI in Science Speaker Series

Organized by the University of Chicago’s Eric and Wendy Schmidt AI in Science Fellowship Program.

Agenda
4:00pm – 4:45pm:  Presentation
4:45pm – 5:00pm:  Q&A
5:00pm – 5:30pm: Reception

Meeting location
William Eckhardt Research Center. Room 401
5640 S Ellis Avenue, Chicago, IL 60637
Map It

Title: Computational Imaging: Restoration Deep Networks as Implicit Priors

Abstract: Computational imaging problems are often formulated as ill-posed inverse problems, requiring the integration of prior knowledge to recover high-quality images from limited or corrupted measurements. This talk focuses on score-based methods, which approximate the gradient of the log-prior (known as the score function) using pre-trained deep restoration networks. I will present recent advances for learning score functions without access to clean training data. These include characterizing score functions via general restoration networks, learning from partial measurements, and model-based strategies that incorporate forward models into training. The talk will also cover the theoretical foundations of these approaches and their applications in biomedical image reconstruction.

Bio: Ulugbek S. Kamilov is the Donald L. Snyder Associate Professor of Electrical & System Engineering and Computer Science & Engineering at WashU, where he founded and leads the Computational Imaging Group (CIG). He received his BSc/MSc in Communication Systems (2011) and PhD in Electrical Engineering (2015) from EPFL, Switzerland. His other prior appointments include Visiting Professor at the École Normale Supérieure in Paris (2024), Visiting Faculty Researcher at Google (2023–2024), and Research Scientist at Mitsubishi Electric Research Laboratories (MERL) (2015-2017).

He is a recipient of the IEEE Signal Processing Society’s 2024 Pierre-Simon Laplace Early Career Technical Achievement Award, the 2017 Best Paper Award, and the NSF CAREER Award. He was named a Scialog Fellow for Advancing Bioimaging in 2021 and was a finalist for the EPFL Doctorate Award in 2016. In 2023, he received the Outstanding Teaching Award from WashU’s Department of Electrical & Systems Engineering. He currently serves on the IEEE Signal Processing Society’s Bioimaging and Signal Processing Technical Committee, and has previously served as a Senior Editorial Board Member of IEEE Signal Processing Magazine, Associate Editor of IEEE Transactions on Computational Imaging, and member of the Computational Imaging Technical Committee.