[2007.13640] Solving Linear Inverse Problems Using the Prior Implicit in a Denoiser

Prior probability models are a fundamental component of many image processing problems, but density estimation is notoriously difficult for high-dimensional signals such as photographic images. Deep neural networks have provided state-of-the-art solutions for problems such as denoising, which implicitly rely on a prior probability model of natural images. Here, we develop a robust and general methodology for making use of this implicit prior. We rely on a statistical result due to Miyasawa (1961), who showed that the least-squares solution for removing additive Gaussian noise can be written directly in terms of the gradient of the log of the noisy signal density. We use this fact to develop a stochastic coarse-to-fine gradient ascent procedure for drawing high-probability samples from the implicit prior embedded within a CNN trained to perform blind (i.e., with unknown noise level) least-squares denoising. A generalization of this algorithm to constrained sampling provides a method for using the implicit prior to solve any linear inverse problem, with no additional training. We demonstrate this general form of transfer learning in multiple applications, using the same algorithm to produce state-of-the-art levels of unsupervised performance for deblurring, super-resolution, inpainting, and compressive sensing.

Prior probability models are a fundamental component of many image processing problems, but density estimation is notoriously difficult for high-dimensional signals such as photographic images. Deep neural networks have provided state-of-the-art solutions for problems such as denoising, which implicitly rely on a prior probability model of natural images. Here, we develop a robust and general methodology for making use of this implicit prior. We rely on a statistical result due to Miyasawa (1961), who showed that the least-squares solution for removing additive Gaussian noise can be written di

saved by

• 3475

related reading

• Unsupervised Learning with SURE.pdfarxiv.org
• Extending Stein's unbiased risk estimator to train deep denoisers with correlated pairs of noisy imagesproceedings.neurips.cc
• [2304.11751] Score-Based Diffusion Models as Principled Priors for Inverse Imagingarxiv.org
• Noiser2Noise.pdfarxiv.org
• Denoising Density Estimatorsarxiv.org
• Deep Energy Estimator.pdfarxiv.org
• Neighbor2Neighborarxiv.org
• Beyond Gaussian Noise: A Generalized Approach to Likelihood Analysis with Non-Gaussian Noiseiopscience.iop.org
• Learning to be Bayesian without Supervisionpapers.nips.cc
• Yang Songyang-song.net
• The Little Book of Deep Learningfleuret.org
• Feature Visualizationdistill.pub