Blind Deblurring Using Internal Patch Recurrence

By Tomer Michaeli, Michal Irani. 2014

437 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

Citations in the corpus

Examining the Impact of Blur on Recognition by Convolutional Networks (2016, background): “Graphics and vision researchers have made significant progress on this problem across the last decade [7, 13, 15, 28]—including with recent neural network-based methods [1, 21, 22, 26].”
SinGAN: Learning a Generative Model From a Single Natural Image (2019, background): “Classical examples include denoising [65], deblurring [39], super resolution [18], dehazing [2, 15], and image editing [37, 21, 9, 11, 50].”
A Neural Approach to Blind Motion Deblurring (2016, methodology): “This is not entirely surprising, since many iterative algorithms (including [2,3]) also divide the inference task into sequential coarse-to-fine reasoning at individual scales.”
Learning to Maintain Natural Image Statistics (2018, background): “This has been shown effective for a range of applications [17,18,19,20].”
A Comparative Study for Single Image Blind Deblurring (2016, methodology): “The progress in this field can be attributed to the advancement of efficient inference algorithms [2] [5] [17] [35] [44], various natural image priors [15] [21] [29] [38] [45], and more general motion blur models [8] [9] [11] [43].”
Image Deblurring via Extreme Channels Prior (2017, background): “Such as heavy-tailed gradient distributions [5, 13], normalized sparsity prior [12], L0-regularized gradient [25], patch recurrence prior [15], and a combination of the intensity and gradient prior [16].”
Neural Blind Deconvolution Using Deep Priors (2019, methodology): “Other speciﬁcally designed regularizations, e.g ., ℓ 1 /ℓ 2 -norm [15], patch-based prior [24, 41], low-rank prior [34] and dark channel prior [29, 50] have also been proposed to identify and preserve salient edges for beneﬁting blur kernel estimation.”
Rethinking Coarse-to-Fine Approach in Single Image Deblurring (2021, background): “It has been demonstrated that different levels of blur in images can be better handled from multi-scale images [19, 18].”
Deep Video Deblurring for Hand-Held Cameras (2017, methodology): “In recent years many successful methods have been introduced [3, 8, 22, 32, 39, 42, 51, 52], see [47] for a recent survey.”
Deep Image Deblurring: A Survey (2022, background): “Early blind deblurring methods focus on removing uniform blur (Cho and Lee 2009; Fergus et al. 2006; Michaeli and Irani 2014; Schuler et al. 2015; Xu and Jia 2010).”
Zero-Shot Super-Resolution Using Deep Internal Learning (2017, background): “This formed the basis for many unsupervised image enhancement methods, including unsupervised SR [4, 5, 6], Blind-SR [14] (when the downscaling kernel is unknown), Blind-Deblurring [15, 1], Blind-Dehazing [2], and more.”
Conditioned Regression Models for Non-blind Single Image Super-Resolution (2015, background): “, [9, 18, 19]) are typically slow and take several minutes per image, making them impractical to build a large corpus of training data.”
Learning to Super-Resolve Blurry Face and Text Images (2017, methodology): “Several recent methods introduce new image priors that favor clear images over blurred ones in the MAP framework [20, 45, 30, 33, 46].”
Learning to Deblur (2014, background): “and Irani [35], where however [27] and [35] have runtimes”
Human-Aware Motion Deblurring (2019, background): “Blind-Image Deblurring: For the uniform blur problem, conventional methods typically resort to natural image priors to estimate latent images [8, 20, 34, 26, 52, 5, 39, 51, 33].”
Positional Encoding as Spatial Inductive Bias in GANs (2020, background): “, stretching objects with the main structures retained or with reasonable texture patch recurrence [25, 33].”