Nonlocally Centralized Sparse Representation for Image Restoration
- W. Dong, Lei Zhang, Guangming Shi, Xin Li
- Computer ScienceIEEE Transactions on Image Processing
- 1 April 2013
The so-called nonlocally centralized sparse representation (NCSR) model is as simple as the standard sparse representation model, and the extensive experiments validate the generality and state-of-the-art performance of the proposed NCSR algorithm.
Image Deblurring and Super-Resolution by Adaptive Sparse Domain Selection and Adaptive Regularization
- W. Dong, Lei Zhang, Guangming Shi, Xiaolin Wu
- Computer ScienceIEEE Transactions on Image Processing
- 6 December 2010
Extensive experiments on image deblurring and super-resolution validate that by using adaptive sparse domain selection and adaptive regularization, the proposed method achieves much better results than many state-of-the-art algorithms in terms of both PSNR and visual perception.
Two-stage image denoising by principal component analysis with local pixel grouping
- Lei Zhang, W. Dong, D. Zhang, Guangming Shi
- EngineeringPattern Recognition
- 1 April 2010
Nonlocal Image Restoration With Bilateral Variance Estimation: A Low-Rank Approach
- W. Dong, Guangming Shi, Xin Li
- Computer ScienceIEEE Transactions on Image Processing
- 1 February 2013
This paper takes a low-rank approach toward SSC and provides a conceptually simple interpretation from a bilateral variance estimation perspective, namely that singular-value decomposition of similar packed patches can be viewed as pooling both local and nonlocal information for estimating signal variances.
Compressive Sensing via Nonlocal Low-Rank Regularization
- W. Dong, Guangming Shi, Xin Li, Yi Ma, F. Huang
- Computer ScienceIEEE Transactions on Image Processing
- 6 June 2014
A nonlocal low-rank regularization approach toward exploiting structured sparsity and its application into CS of both photographic and MRI images is proposed and the use of a nonconvex log det as a smooth surrogate function for the rank instead of the convex nuclear norm is proposed.
Sparse Representation Based Image Interpolation With Nonlocal Autoregressive Modeling
- W. Dong, Lei Zhang, R. Lukac, Guangming Shi
- Computer ScienceIEEE Transactions on Image Processing
- 1 April 2013
This paper incorporates the image nonlocal self-similarity into SRM for image interpolation, and shows that the NARM-induced sampling matrix is less coherent with the representation dictionary, and consequently makes SRM more effective forimage interpolation.
Hyperspectral Image Super-Resolution via Non-Negative Structured Sparse Representation
- W. Dong, Fazuo Fu, Xin Li
- Environmental Science, Computer ScienceIEEE Transactions on Image Processing
- 22 March 2016
A new hyperspectral image super-resolution method from a low-resolution (LR) image and a HR reference image of the same scene to improve the accuracy of non-negative sparse coding and to exploit the spatial correlation among the learned sparse codes.
Sparsity-based image denoising via dictionary learning and structural clustering
- W. Dong, Xin Li, Lei Zhang, Guangming Shi
- Computer ScienceComputer Vision and Pattern Recognition
- 20 June 2011
A double-header l1-optimization problem where the regularization involves both dictionary learning and structural structuring is formulated and a new denoising algorithm built upon clustering-based sparse representation (CSR) is proposed.
Centralized sparse representation for image restoration
- W. Dong, Lei Zhang, Guangming Shi
- Computer ScienceVision
- 6 November 2011
A novel sparse representation model called centralized sparse representation (CSR) is proposed, which achieves convincing improvement over previous state-of-the-art methods on image restoration tasks by exploiting the nonlocal image statistics.
Denoising Prior Driven Deep Neural Network for Image Restoration
- W. Dong, Peiyao Wang, W. Yin, Guangming Shi, Fangfang Wu, Xiaotong Lu
- Computer ScienceIEEE Transactions on Pattern Analysis and Machine…
- 21 January 2018
A convolutional neural network (CNN) based denoiser that can exploit the multi-scale redundancies of natural images is proposed, which not only exploits the powerful denoising ability of DNNs, but also leverages the prior of the observation model.
...
...