BM3D Image Denoising with Shape-Adaptive Principal Component Analysis

By Kostadin Dabov, Alessandro Foi, Vladimir Katkovnik, Karen O. Egiazarian. 2009

524 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/52bba8a6f6ac79796d8d93d012197d22def599ff

How researchers see this paper

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

Citations in the corpus

A Nonlocal Bayesian Image Denoising Algorithm (2013, methodology): “Nevertheless, the variant with shape-adaptive patches BM3D-SAPCA [14] is sharper but works only on grey level images.”
An Analysis and Implementation of the BM3D Image Denoising Method (2012, methodology): “Like for NL-means, there is a variant with shape-adaptive patches [5].”
Nonlocally Centralized Sparse Representation for Image Restoration (2013, methodology): “…NCSR method with three recently developed state-of-the-art denoising methods, including the shape-adaptive PCA based BM3D (SAPCA-BM3D) [39], [40] (which outperforms the benchmark BM3D algorithm [17]), the learned simultaneously sparse coding (LSSC) method [18] and the expected patch log…”
Brief review of image denoising techniques (2019, background): “To improve denoising performance, many improved versions of BM3D have appeared [97, 98].”
Image Noise Level Estimation by Principal Component Analysis (2013, methodology): “It is based on principal component analysis (PCA) of image blocks, which has been already successfully utilized in various image processing tasks such as compression [33], denoising [34]–[36], and quality assessment [37].”
One Network to Solve Them All — Solving Linear Inverse Problems Using Deep Projection Models (2017, methodology): “For image denoising, we compare with the state-of-the-art algorithm, BM3D [38].”
Image Denoising Using the Higher Order Singular Value Decomposition (2013, unknown): “But we have observed that it falls short of the shape adaptive implementation of BM3D [10].”
Real Image Denoising With Feature Attention (2019, methodology): “NLNet improved upon the traditional methods; but, it lags in performance compared to most of the CNNs [60, 59] due to the adaptaton of NSS priors, as it is unable to ﬁnd the analogs for all the patches in the image.”
Unsupervised Image Super-Resolution Using Cycle-in-Cycle Generative Adversarial Networks (2018, background): “Although BM3D can remove noise, it also over-smooth the input images.”
Under Review as a Conference Paper at Iclr 2016 Predicting Distributions with Linearizing Belief Networks (2015, methodology): “Previous approaches such as the state-of-the-art BM3D (Dabov et al., 2009) method ignore this difficulty and simply predict the conditional average.”
GAN-Based Projector for Faster Recovery With Convergence Guarantees in Linear Inverse Problems (2019, background): “Traditionally, for signal priors, natural images are considered sparse in some fixed or learnable basis [11, 8, 36, 22, 7, 38, 10, 21].”
From Denoising to Compressed Sensing (2014, methodology): “Additionally, the packages provided with many of the state-of-the-art denoising algorithms [55], [61], [74], work with just two inputs; the noisy signal and an estimate of the standard deviation of the Gaussian noise.”
Iterative Weighted Maximum Likelihood Denoising With Probabilistic Patch-Based Weights (2009, background): “The best denoising techniques so far for additive Gaussian noise [26]–[28] consist of a combination of the ideas of similar patches selection with that of sparsification by transforms or learned dictionaries.”
From Local Kernel to Nonlocal Multiple-Model Image Denoising (2009, methodology): “The use of a data-driven adaptive transform for the collaborative Þltering results in a further improvement of the denoising performance, especially in preserving image details and introducing very few artifacts [21].”
Exploiting the Potential of Standard Convolutional Autoencoders for Image Restoration by Evolutionary Search (2018, methodology): “Comparison of results of the BM3D (Dabov et al., 2009), RED (Mao et al., 2016), MemNet (Tai et al., 2017), and E-CAE.”
Poisson Noise Reduction with Non-local PCA (2012, methodology): “Since the introduction of non-local methods for image denoising [8], these methods have proved to outperform previously considered approaches [1, 11, 12, 30] (extensive comparisons of recent denoising method can be found for Gaussian noise in [21, 26]).”