Join

We are always looking for curious, motivated people to work on computational imaging.

PhD StudentsRecruiting for Fall 2027

The Computational Imaging Group (CIG) at the University of Wisconsin–Madison is recruiting a PhD student for Fall 2027. Our current research interests include generative models, self-supervised learning, optimization, and large-scale inference, with applications to computational imaging and inverse problems. We combine physical models with machine learning to reconstruct and understand the physical world from incomplete and noisy measurements.

CIG recruits students through both Electrical and Computer Engineering (ECE) and Computer Sciences (CS). Applicants should apply through the corresponding graduate program and mention Ulugbek Kamilov in their statement of purpose.

Apply →

Student ResearchersRecruiting

UW–Madison BS and MS students can work on research projects in CIG. Projects are co-mentored by a PhD students or postdocs and Prof. Kamilov. Projects usually run for one or two semesters and usually are done for credit.

Apply →

Reserved · Yan LiFall 2026

Generative Models as Priors for Ptychography

Mentors: Edward Chandler, Ulugbek S. Kamilov

Ptychography reconstructs high-resolution images from overlapping diffraction measurements, but limited or noisy data can make reconstruction challenging. This project explores generative models as learned priors to guide reconstruction toward images consistent with both the measurements and the structure of the sample. By combining physical models of diffraction with generative modeling and optimization, we aim to improve reconstruction quality and reduce data acquisition requirements.

Reserved · Joshua WangFall 2026

Motion Correction in 3D Whole-Body MRI

Mentors: Chicago Park, Ulugbek S. Kamilov

Patient motion during 3D whole-body MRI can blur images and obscure anatomical details. In collaboration with the UW School of Medicine, this project develops computational methods to estimate and correct motion during image reconstruction. By combining MRI physics with optimization and generative modeling, we aim to improve image quality and the reliability of whole-body imaging.

Reserved · Tyler SeilsFall 2026

Reconstructing MRI Images with Diffusion Models

Mentors: Xiao Shi, Ulugbek S. Kamilov

This project develops MRI reconstruction methods that use diffusion models as learned image priors. By combining generative modeling with MRI physics, we aim to recover high-quality images from undersampled measurements, enabling faster scans. The work is conducted in collaboration with the UW School of Medicine.