# DeepRED: Deep Image Prior Powered by RED Source: [DeepRED: Deep Image Prior Powered by RED]() Collection index: [llms.txt]() Data updated: 2026-01-04T04:08:26.260Z Authors: Gary Mataev, Michael Elad, Peyman Milanfar Published: 2019-03-25T08:39:53.000Z ## Abstract Inverse problems in imaging are extensively studied, with a variety of strategies, tools, and theory that have been accumulated over the years. Recently, this field has been immensely influenced by the emergence of deep-learning techniques. One such contribution, which is the focus of this paper, is the Deep Image Prior (DIP) work by Ulyanov, Vedaldi, and Lempitsky (2018). DIP offers a new approach towards the regularization of inverse problems, obtained by forcing the recovered image to be synthesized from a given deep architecture. While DIP has been shown to be quite an effective unsupervised approach, its results still fall short when compared to state-of-the-art alternatives. In this work, we aim to boost DIP by adding an explicit prior, which enriches the overall regularization effect in order to lead to better-recovered images. More specifically, we propose to bring-in the concept of Regularization by Denoising (RED), which leverages existing denoisers for regularizing inverse problems. Our work shows how the two (DIP and RED) can be merged into a highly effective unsupervised recovery process while avoiding the need to differentiate the chosen denoiser, and leading to very effective results, demonstrated for several tested problems. ## Recorded benchmark results A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. - [Image Deblurring on Set5 Gaussian Blur (test)](): Image Deblurring; Set5 Gaussian Blur (test); Baby Score: 35.3. - [Color Image Deblurring on Butterfly, Leaves, Parrots, Starfish Uniform Blur](): Color Image Deblurring; Butterfly, Leaves, Parrots, Starfish Uniform Blur; Quality Score (Butterfly): 31.44. - [Color Image Deblurring on Butterfly, Leaves, Parrots, Starfish Gaussian Blur](): Color Image Deblurring; Butterfly, Leaves, Parrots, Starfish Gaussian Blur; Score (Butterfly): 32.19. - [Image Deblurring on Set5 Uniform Blur (test)](): Image Deblurring; Set5 Uniform Blur (test); Baby Category Score: 33.11. - [Single Image Super-Resolution on Set5 8:1 (test)](): Single Image Super-Resolution; Set5 8:1 (test); PSNR: 26.04. - [Single Image Super-Resolution on Set14 8:1 (test)](): Single Image Super-Resolution; Set14 8:1 (test); PSNR: 24.28. - [Single Image Super-Resolution on Set14 (Per-Image Scores)](): Single Image Super-Resolution; Set14; baboon: 22.51. - [Single Image Super-Resolution on Set5 4:1 (test)](): Single Image Super-Resolution; Set5 4:1 (test); PSNR: 30.72. - [Super-Resolution on Set14 4x (test)](): Super-Resolution; Set14 4x (test); PSNR: 27.63. - [Single Image Super-Resolution on Set14 (PSNR, SSIM)](): Single Image Super-Resolution; Set14; PSNR: 24.28. ## Original sources and implementation [Original paper]() [PDF]()