News

Welcome to Harry Mellow, who has joined CIG as a PhD student.

Stochastic Generative Plug-and-Play Priors was accepted to IEEE Transactions on Computational Imaging.

New paper SNaP: One-Step Posterior Sampling for Noisy Inverse Problems.

Prof. Kamilov was a plenary speaker at IEEE ICIP 2026, which took place in Tampere, Finland.

New paper FIRM: Flow-based Imaging via Regularized Minimization.

New paper Next-Dense-Stride Prediction for Multimodal Autoregressive Visual Modeling.

Prof. Kamilov gave a talk at the AI for Multimodal Science workshop in San Diego, CA.

New paper NullFlow: One-Step Generative Reconstruction.

RELD: Regularization by Latent Denoising was accepted to IEEE Signal Processing Letters.

Prof. Kamilov spoke at the IEEE SPS Webinar.

Three summer interns join CIG: Sam Cook, Joshua Wang, and Tyler Seils.

Prof. Kamilov has been appointed an Affiliate Professor of Radiology at UW–Madison.

Prof. Kamilov has been appointed an Affiliate Professor of Computer Sciences at UW–Madison.

Prof. Kamilov was a guest on the Russian-language SEREDIN podcast.

Prof. Kamilov has become an affiliate of the Data Science Institute at UW–Madison.

AI-Powered Gradient Echo Plural Contrast Imaging (AI-GEPCI)—A Comprehensive Neurological Protocol From a Single MRI Scan was accepted to Journal of Magnetic Resonance Imaging.

New paper Stochastic Generative Plug-and-Play Priors.

Prof. Kamilov is serving as an area chair for NeurIPS 2026.

Prof. Kamilov spoke at the Machine Learning for Medical Imaging seminar series at UW–Madison.

New paper A Unified Framework for Multimodal Image Reconstruction and Synthesis using Denoising Diffusion Models.

Measurement Score-Based Diffusion Model was accepted to ICLR 2026.

EigenScore: OOD Detection using Posterior Covariance in Diffusion Models was accepted to ICLR 2026.

Prof. Kamilov is teaching Matrix Methods in Machine Learning in Spring 2026 at UW–Madison.

Welcome to Xiao Shi, who has joined CIG as a PhD student.

Moments Matter: Posterior Recovery in Poisson Denoising via Log-Networks was accepted to ICASSP 2026.

New paper Deep Parameter Interpolation for Scalar Conditioning.

Prof. Kamilov served as an area chair at CVPR 2026.

Prof. Kamilov served as an area chair at ICML 2026.

Prof. Kamilov started as Leon and Elizabeth Janssen Associate Professor of ECE at UW–Madison.

Prof. Kamilov gave a talk at the Schmidt AI in Science Speaker Series on October 21, 2025, at the University of Chicago.

Prof. Kamilov gave a talk at the Problèmes inverses (aveugles) en imagerie: des fondements aux applications on September 29 - October 3, 2025, at the CIRM in Marseille, France.

Kernel Density Steering: Inference-Time Scaling via Mode Seeking for Image Restoration was accepted to NeurIPS 2025.

Distribution Shift Estimation in Imaging Inverse Problems was accepted to IEEE CAMSAP 2025.

Closed-Form Approximation of the Total Variation Proximal Operator was accepted to IEEE Transactions on Computational Imaging.

A Generalizable Diffusion Framework for 3D Low-Dose and Few-View Cardiac SPECT Imaging was accepted to Medical Image Analysis.

Computational field-resolved coherent chemical imaging was accepted to Nature Communications.

Gaussian is All You Need: A Unified Framework for Solving Inverse Problems via Diffusion Posterior Sampling was accepted to IEEE Transactions on Computational Imaging.

Advanced Quantification Pipeline Reveals New Spatial and Temporal Tumor Characteristics in Preclinical Multiple Myeloma was accepted to EJNMMI Research.

Random Walks with Tweedie: A Unified View of Score-Based Diffusion Models was accepted to IEEE Signal Processing Magazine.

CIG PhD student Weijie Gan has successfully defended his PhD thesis Computational Imaging under Incomplete Information. Congratulations to Weijie.

New paper Unsupervised Detection of Distribution Shift in Inverse Problems using Diffusion Models. Collaboration with Salman Asif from UC Riverside.

New paper Measurement Score-Based Diffusion Model. Collaboration with Hongyu An from WashU Medicine.

Plug-and-Play Priors as a Score-Based Method was accepted to IEEE ICIP 2025. Collaboration with Mike McCann, Christina Garcia-Cardona, and Brendt Wohlberg from Los Alamos National Lab.

Stochastic Deep Restoration Priors for Imaging Inverse Problems was accepted to ICML 2025. Collaboration with Mauricio Delbracio and Peyman Milanfar from Google.

Former CIG PhD student Jiaming Liu was awarded the Outstanding Doctoral Dissertation Award by the Department of Electrical and Systems Engineering at WashU.

Prof. Kamilov is the inaugural recipient of the Donald L. Snyder Career Development Professorship at WashU.

Prof. Kamilov received the IEEE SPS Pierre-Simon Laplace Early Career Technical Achievement Award at IEEE ICASSP 2025 “for major contributions to theory and practice in computational imaging”.

Prof. Kamilov gave a talk on Computational imaging: Restoration deep networks as implicit priors at the Deep Reconstruction Workshop that took place at the Johns Hopkins University (see the recording here).

Efficient Model-based Deep Learning via Network Pruning and Fine-Tuning was accepted to Journal of Mathematical Imaging and Vision.

CoRRECT: A Deep Unfolding Framework for Motion-Corrected Quantitative R2* Mapping was accepted to Journal of Mathematical Imaging and Vision.

FiRe: Fixed-points of Restoration Priors for Solving Inverse Problems was accepted to IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

CDDIP: Constrained Diffusion-Driven Deep Image Prior for Seismic Data Reconstruction was accepted to IEEE Geoscience and Remote Sensing Letters.

Pseudo-MRI-Guided PET Image Reconstruction Method Based on a Diffusion Probabilistic Model was accepted to IEEE Transactions on Radiation and Plasma Medical Sciences.

A Self-Supervised Diffusion Bridge for MRI Reconstruction was accepted to IEEE International Symposium on Biomedical Imaging (ISBI).

New paper Plug-and-Play Priors as a Score-Based Method. Collaboration with Michael McCann, Cristina Garcia-Cardona, and Brendt Wohlberg from LANL.

New paper Closed-Form Approximation of the Total Variation Proximal Operator. Collaboration with Brendt Wohlberg from LANL.

New paper Random Walks with Tweedie: A Unified Framework for Diffusion Models. Collaboration with Michael McCann, Cristina Garcia-Cardona, and Brendt Wohlberg from LANL.

New paper FiRe: Fixed-points of Restoration Priors for Solving Inverse Problems. Collaboration with Matthieu Terris and Thomas Moreau from INRIA Saclay.

New paper ADOBI: Adaptive Diffusion Bridge For Blind Inverse Problems with Application to MRI Reconstruction.

Fast motion-compensated reconstruction for 4D-CBCT using deep learning-based groupwise registration was accepted to Biomedical Physics & Engineering Express.

New paper Stochastic Deep Restoration Priors for Imaging Inverse Problems. This paper was done in collaboration with Mauricio Delbracio and Peyman Milanfar from Google.

Chicago Park joins the Computational Imaging Group (CIG) as a PhD student. He was previously a research intern at the Los Alamos National Laboratory. He graduated summa cum laude with a BS degree in Computer Science from WashU.

FLAIR: A Conditional Diffusion Framework with Applications to Face Video Restoration was accepted to IEEE/CVF Winter Conference on Applications of Computer Vision (WACV).

Bioengineering approaches for patient-specific analysis of placenta structure and function was accepted to Placenta.

Prof. Kamilov co-organized the IMSI Workshop on Computational Imaging in Chicago, IL, USA, on 5-9 August.

Former CIG PhD student Yu Sun is now an Assistant Professor at the ECE department of the Johns Hopkins University.

Prof. Kamilov received an unrestricted research gift from Google Research for the project “Innovative algorithms for computing solutions to imaging inverse problems using pretrained (generative) models”.

New paper CDDIP: Constrained Diffusion-Driven Deep Image Prior for Seismic Image Reconstruction.

Prof. Kamilov completed a sabbatical at the École Normale Supérieure in Paris, France. Many thanks to the hosts, Sylvain Gigan and Giulio Biroli, as well as to Gabriel Peyré.

Former CIG PhD student Xiaojian Xu is now an AI Research Scientist at GE HealthCare.

CIG PhD student Jiaming Liu has defended his PhD thesis “Learning-based Artifacts Removing Priors for Computational Imaging”. He is now a Postdoctoral Fellow at Stanford.