Deep Image Prior

Latest papers

9 papers1 paper in the last four weeks, against 1 the four weeks before. 0.0% of all new papers.

Neural Blind Deconvolution for Image Scanning Microscopy
Oct 29, 2026Eliot Guez, Nathan Buskulic, Lisa Cuneo +2Microscopy Image Analysis Image Super-Resolution
MaLGa Center, DIBRIS, University of Genoa, Italy · CSLM, Italian Institute of Technology, Italy · MMS, Italian Institute of Technology, Italy
Interpretable Hyperspectral Unmixing Framework with Fixed Endmember Prior and Structured Residual Refinement
Sep 8, 2026Ziyi Guan, Jianping Zhang, Qian LiuDeep Image Prior Hyperspectral Imaging
School of Mathematics and Computational Science, Xiangtan University, Xiangtan 411105, China
Pooling-Based Context Modeling for Convolution-Free Deep Image Prior
Jul 3, 2026Gihyun Kim, Jong-Seok LeeDeep Image Prior Image Restoration
Yonsei University Republic of Korea
A Zero-Shot Deep Image Prior Framework for Denoising and Deconvolution in Fluorescence Microscopy
Jun 26, 2026Xiangyu Qian, Jing Liu, Yunqing Tang +2Deep Image Prior Physics-Guided Image Restoration
Wenzhou Key Laboratory of Biomedical Imaging, Zhejiang Key Laboratory of Soft Matter Biomedical Materials, Wenzhou Institute, University of Chinese Academy of Sciences, Wenzhou, Zhejiang 325000, China · Postgraduate Training Base Alliance of Wenzhou Medical University, Wenzhou, Zhejiang 325035, China
Unsupervised Deep Learning for Limited-Angle STEM-EDX Tomography -- Application to 3D Chemical Analysis of Phase-Change Memory Devices
Jun 9, 2026Daniel del Pozo Bueno, Serge Brosset, Theo Monniez +3Deep Image Prior Image Reconstruction
1CEA, LETI, Univ. Grenoble Alpes, Grenoble, 38000, France. · 2CEA, Neurospin, Paris-Saclay University, Gif-sur-Yvette, F-91191, France. · 3Inria, MIND, Palaiseau, F-91210, France.
Unsupervised Deep Image Prior for Sparse-View and Limited-Angle Electron Tomography
May 26, 2026Serge Brosset, Daniel del Pozo Bueno, Thomas David +3Deep Image Prior Sparse-View CT Reconstruction
a Univ. Grenoble Alpes, CEA, Leti, F-38000 Grenoble, France. · b Univ. Grenoble Alpes, CEA, Liten, F-38000 Grenoble, France. · c CEA, Joliot, NeuroSpin, Gif-sur-Yvette Cedex 91191, France. +1
A Principled Self-Referenced Early Stopping Approach for Deep Image Prior
May 24, 2026Chaoyan Huang, Cheng-Han Huang, Ismail R. Alkhouri +1Deep Image Prior Early Stopping
Department of Computational Mathematics, Science, & Engineering, Michigan State University · Department of Electrical Engineering and Computer Science, University of Michigan · X Computational Physics Division, Los Alamos National Laboratory +2
Deep Image Prior for photoacoustic tomography can mitigate limited-view artifacts
Apr 21, 2026Hanna Pulkkinen, Jenni Poimala, Leonid Kunyansky +2Deep Image Prior Medical Imaging
1Research Unit of Mathematical Sciences, University of Oulu, Finland · Department of Technical Physics, University of Eastern Finland, Finland · Department of Mathematics, University of Arizona, USA +2
Preventing Overfitting in Deep Image Prior for Hyperspectral Image Denoising
Apr 9, 2026Panagiotis Gkotsis, Athanasios A. RontogiannisDeep Image Prior Image Denoising
Robotics Institute, Athena Research and Innovation Center, 151 25 Maroussi, Greece · School of Electrical and Computer Engineering, National Technical University of Athens, 157 80 Athens, Greece

Oct 29, 2026J/K move · Enter open · S save

Neural Blind Deconvolution for Image Scanning Microscopy

Eliot Guez, Nathan Buskulic, Lisa Cuneo, Luca Calatroni+1

MaLGa Center, DIBRIS, University of Genoa, Italy · CSLM, Italian Institute of Technology, Italy · MMS, Italian Institute of Technology, Italy

Microscopy Image Analysis Image Super-Resolution Image Deblurring Deep Image Prior

Open paper AskPDF arXiv

Image scanning microscopy (ISM) acquires one scanned image per detector element, which are then fused into a single image by multi-image deconvolution to obtain a super-resolved image. Its accuracy, however, depends on point spread function (PSF) models that sample- and system-induced aberrations often invalidate. Since differences between the scanned images encode aberration information, we propose a training-free blind Deep Image Prior (DIP) framework that jointly estimates the specimen and the detector-dependent PSFs. The PSFs are parametrised through the detector geometry and shared excitation and emission PSFs, reducing the unknowns while preserving the physical ISM model, and a Gaussian pre-initialisation stabilises the blind optimisation. On experimental tubulin data, blind DIP recovers finer filamentous structures than multi-image deconvolution and DIP with fixed theoretical PSFs, while adapting the PSFs to the measurements.