Random cascades of Gaussian scale mixtures and their use in modeling natural images with application to denoising
Published 11 November 2002
Martin J. Wainwright, Eero P. Simoncelli, Alan S. Willsky
Citations16
Generate an AI Snapshot to get a quick, structured summary of this paper.
Study Snapshot
ObjectiveStudy objective
MethodsResearch methodology
PopulationPopulation studied
Sample sizeSample sizes
OutcomesStudy outcomes here
ResultsStudy results comes here
LimitationsResearch study limitations comes here
A concise AI-generated summary of the paper will appear here once you click Generate AI Snapshot.
TL;DR
A Newton-like method for exact MAP state estimation that exploits fast algorithms for tree estimation, and hence is very efficient and related to a number of previous approaches to image coding and denoising are developed.
Abstract
Multiresolution representations play an important role in image processing and computer vision, as well as in modeling stochastic processes. We have developed a semi-parametric class of non-Gaussian multiscale statistical processes defined by random cascades on wavelet trees. This model class is rich enough to accurately capture the remarkably regular non-Gaussian features of natural images, but sufficiently structured to permit estimation of the underlying state variables. We showed that our models accurately fit both the marginal and joint histograms of wavelet coefficients from natural images. We developed a Newton-like method for exact MAP state estimation that exploits fast algorithms for tree estimation, and hence is very efficient. Applications of this algorithm to denoising of both 1D signals and natural images were presented. The GSM-tree model class is related to a number of previous approaches to image coding and denoising.
Keywords
Computer ScienceEngineering
IEEE Transactions on Signal ProcessingEmbedded image coding using zerotrees of wavelet coefficients
4,813 Citations1993J.M. Shapiro
Journal of the Optical Society of America ARelations between the statistics of natural images and the response properties of cortical cells
3,233 Citations1987David Field
The results obtained with six natural images suggest that the orientation and the spatial-frequency tuning of mammalian simple cells are well suited for coding the information in such images if the goal of the code is to convert higher-order redundancy into first- order redundancy.
Proceedings - International Conference on Image ProcessingThe steerable pyramid: a flexible architecture for multi-scale derivative computation
1,062 Citations2002Eero P. Simoncelli, William T. Freeman
An architecture for efficient and accurate linear decomposition of an image into scale and orientation subbands and the construction and implementation of the transform is described.
IEEE Signal Processing LettersLow-complexity image denoising based on statistical modeling of wavelet coefficients
777 Citations1999M. Kıvanç Mıhçak, I. Kozintsev +2 more
This work introduces a simple spatially adaptive statistical model for wavelet image coefficients and applies it to image denoising, inspired by a recentWavelet image compression algorithm, the estimation-quantization (EQ) coder.
IEEE Transactions on Image ProcessingImage compression via joint statistical characterization in the wavelet domain
522 Citations1999R.W. Buccigrossi, Eero P. Simoncelli
A Markov model is proposed that explains dependencies using a linear predictor for magnitude coupled with both multiplicative and additive uncertainties, and it is shown that it accounts for the statistics of a wide variety of images including photographic images, graphical images, and medical images.
Statistics of natural images and models
510 Citations2003Jinggang Huang, D. Mumford
Results ranging from the simplest single pixel intensity to joint distribution of 3 Haar wavelet responses are reported, which shed light on old issues such as the near scale-invariance of image statistics and some are entirely new.
Scale Mixtures of Gaussians and the Statistics of Natural Images
442 Citations1999Martin J. Wainwright, Eero P. Simoncelli
Proceedings of the IEEEWavelet-based representations for the 1/f family of fractal processes
322 Citations1993Gregory W. Wornell
IEEE Transactions on Information TheoryCorrelation structure of the discrete wavelet coefficients of fractional Brownian motion
297 Citations1992Ahmed H. Tewfik, M. Kim
It is shown that the discrete wavelet coefficients of fractional Brownian motion at different scales are correlated and that their auto- and cross-correlation functions decay hyperbolically fast at a rate much faster than that of the autocorrelation of the fractionalBrownian motion itself.
IEEE Transactions on Automatic ControlMultiscale recursive estimation, data fusion, and regularization
281 Citations1994K.C. Chou, Alan S. Willsky +1 more
The main focus of this paper is on the description, analysis, and application of an extremely efficient optimal estimation algorithm for this class of multiscale dynamic models evolving on dyadic trees.
Statistical models for images: compression, restoration and synthesis
270 Citations2002Eero P. Simoncelli
A parametric statistical model for visual images in the wavelet transform domain is presented and the joint densities of coefficient magnitudes at adjacent spatial locations, adjacent orientations, and adjacent spatial scales are characterized.
Image coding based on mixture modeling of wavelet coefficients and a fast estimation-quantization framework
258 Citations2002Scott M. LoPresto, Kannan Ramchandran +1 more
A new image compression paradigm that combines compression efficiency with speed, and is based on an independent "infinite" mixture model which accurately captures the space-frequency characterization of the wavelet image representation, is introduced.
Lecture notes in statisticsBayesian Denoising of Visual Images in the Wavelet Domain
240 Citations1999Eero P. Simoncelli
The most well-known description of image statistics is that their Fourier spectra take the form of a power law, which suggests that the Fourier transform is an appropriate PCA representation of Fourier and related representations, widely used in image processing applications.
Applied and Computational Harmonic AnalysisRandom Cascades on Wavelet Trees and Their Use in Analyzing and Modeling Natural Images
200 Citations2001Martin J. Wainwright, Eero P. Simoncelli +1 more
Physical Review LettersSelf-Similarity Properties of Natural Images Resemble Those of Turbulent Flows
82 Citations1998Antonio Turiel, Germán Mato +2 more
Bayesian wavelet-domain image modeling using hidden Markov trees
21 Citations2003Justin Romberg, Hyun-Ho Choi +1 more
Two reduced-parameter HMT models that capture the general structure of a broad class of grayscale images are proposed and a shift-invariant wavelet denoising scheme is developed that outperforms all algorithms in the current literature.
Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIERandom cascades on wavelet trees and their use in analyzing and modeling natural images
4 Citations2000Martin J. Wainwright, Eero P. Simoncelli +1 more
A Newton- like algorithm for MAP estimation that exploits very fast algorithm for linear-Gaussian estimation on trees, and hence is efficient, and a denoising technique that is based on a global prior model, and preserves the structure of natural images is developed.