Image Deconvolution with Deep Image and Kernel Priors

Wang, Zhunxuan; Wang, Zipei; Bilen, Hakan

doi:10.48550/arxiv.1910.08386

Cited by 1 publication

(2 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…BID is a very challenging visual IP due to the bilinearity: (k, x) → k * x. Recently, [2,33,39,42] have tried to use DIP models to solve BID by modeling k and x as two separate DNNs, i.e.,…”

Section: Blind Image Deblurring (Bid)mentioning

confidence: 99%

“…The simple procedure has produced surprisingly competitive recovery results on a myriad of visual IPs, from low-level image denoising, super-resolution, inpainting [14,25,40] and blind deconvolution [2,33,39,42], to mid-level image decomposition [10] and image fusion [26], and to advanced computational imaging problems [3, 4, 8, 11-13, 41, 43, 44] (see the survey [32]). A salient feature of the procedure is no extra training data, despite the presence of DNNs: the G θ 's-often convolutional neural networks (CNNs)-are chosen to simply enforce structural priors for natural visual objects.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Early Stopping for Deep Image Prior

Wang¹,

Li²,

Zhang³

et al. 2021

Preprint

View full text Add to dashboard Cite

Deep image prior (DIP) and its variants have showed remarkable potential for solving inverse problems in computer vision, without any extra training data. Practical DIP models are often substantially overparameterized. During the fitting process, these models learn mostly the desired visual content first, and then pick up the potential modeling and observational noise, i.e., overfitting. Thus, the practicality of DIP often depends critically on good early stopping (ES) that captures the transition period. In this regard, the majority of DIP works for vision tasks only demonstrates the potential of the models-reporting the peak performance against the ground truth, but provides no clue about how to operationally obtain near-peak performance without access to the groundtruth. In this paper, we set to break this practicality barrier of DIP, and propose an efficient ES strategy, which consistently detects near-peak performance across several vision tasks and DIP variants. Based on a simple measure of dispersion of consecutive DIP reconstructions, our ES method not only outpaces the existing ones-which only work in very narrow domains, but also remains effective when combined with a number of methods that try to mitigate the overfitting. The code is available at https://github.com/sunumn/Early_Stopping_for_DIP.

show abstract

Section: Blind Image Deblurring (Bid)mentioning

confidence: 99%