Compact Support Biorthogonal Wavelet Filterbanks for Arbitrary Undirected Graphs
TL;DR: This paper relax the condition of orthogonality to design a biorthogonal pair of graph-wavelets that are k-hop localized with compact spectral spread and still satisfy the perfect reconstruction conditions.
Abstract: This paper extends previous results on wavelet filterbanks for data defined on graphs from the case of orthogonal transforms to more general and flexible biorthogonal transforms. As in the recent work, the construction proceeds in two steps: first we design “one-dimensional” two-channel filterbanks on bipartite graphs, and then extend them to “multi-dimensional” separable two-channel filterbanks for arbitrary graphs via a bipartite subgraph decomposition. We specifically design wavelet filters based on the spectral decomposition of the graph, and state sufficient conditions for the filterbanks to be perfect reconstruction and orthogonal. While our previous designs, referred to as graph-QMF filterbanks, are perfect reconstruction and orthogonal, they are not exactly k-hop localized, i.e., the computation at each node is not localized to a small k-hop neighborhood around the node. In this paper, we relax the condition of orthogonality to design a biorthogonal pair of graph-wavelets that are k-hop localized with compact spectral spread and still satisfy the perfect reconstruction conditions. The design is analogous to the standard Cohen-Daubechies-Feauveau's (CDF) construction of factorizing a maximally-flat Daubechies half-band filter. Preliminary results demonstrate that the proposed filterbanks can be useful for both standard signal processing applications as well as for signals defined on arbitrary graphs.
Citations
More filters
25 Apr 2018
TL;DR: An overview of core ideas in GSP and their connection to conventional digital signal processing are provided, along with a brief historical perspective to highlight how concepts recently developed build on top of prior research in other areas.
Abstract: Research in graph signal processing (GSP) aims to develop tools for processing data defined on irregular graph domains. In this paper, we first provide an overview of core ideas in GSP and their connection to conventional digital signal processing, along with a brief historical perspective to highlight how concepts recently developed in GSP build on top of prior research in other areas. We then summarize recent advances in developing basic GSP tools, including methods for sampling, filtering, or graph learning. Next, we review progress in several application areas using GSP, including processing and analysis of sensor network data, biological data, and applications to image processing and machine learning.
1,306 citations
Cites background from "Compact Support Biorthogonal Wavele..."
...graphs [68], [102], thus requiring the graph to be decomposed into a series of bipartite subgraphs [68], [103]....
[...]
TL;DR: It is shown that perfect recovery is possible for graph signals bandlimited under the graph Fourier transform and the connection to the sampling theory of finite discrete-time signal processing and previous work on signal recovery on graphs is established.
Abstract: We propose a sampling theory for signals that are supported on either directed or undirected graphs. The theory follows the same paradigm as classical sampling theory. We show that perfect recovery is possible for graph signals bandlimited under the graph Fourier transform. The sampled signal coefficients form a new graph signal, whose corresponding graph structure preserves the first-order difference of the original graph signal. For general graphs, an optimal sampling operator based on experimentally designed sampling is proposed to guarantee perfect recovery and robustness to noise; for graphs whose graph Fourier transforms are frames with maximal robustness to erasures as well as for Erdős-Renyi graphs, random sampling leads to perfect recovery with high probability. We further establish the connection to the sampling theory of finite discrete-time signal processing and previous work on signal recovery on graphs. To handle full-band graph signals, we propose a graph filter bank based on sampling theory on graphs. Finally, we apply the proposed sampling theory to semi-supervised classification of online blogs and digit images, where we achieve similar or better performance with fewer labeled samples compared to previous work.
644 citations
Cites background from "Compact Support Biorthogonal Wavele..."
...The framework models that underlying structure by a graph and signals by graph signals, generalizing concepts and tools from classical discrete signal processing to graph signal processing....
[...]
Posted Content•
TL;DR: Graph Signal Processing (GSP) as discussed by the authors aims to develop tools for processing data defined on irregular graph domains, including sampling, filtering, and graph learning, which can be used for processing sensor network data, biological data, and image processing and machine learning.
Abstract: Research in Graph Signal Processing (GSP) aims to develop tools for processing data defined on irregular graph domains. In this paper we first provide an overview of core ideas in GSP and their connection to conventional digital signal processing. We then summarize recent developments in developing basic GSP tools, including methods for sampling, filtering or graph learning. Next, we review progress in several application areas using GSP, including processing and analysis of sensor network data, biological data, and applications to image processing and machine learning. We finish by providing a brief historical perspective to highlight how concepts recently developed in GSP build on top of prior research in other areas.
463 citations
TL;DR: This work introduces quantities called graph spectral proxies, defined using the powers of the variation operator, in order to approximate the spectral content of graph signals, and forms a direct sampling set selection approach that does not require the computation and storage of the basis elements.
Abstract: We study the problem of selecting the best sampling set for bandlimited reconstruction of signals on graphs. A frequency domain representation for graph signals can be defined using the eigenvectors and eigenvalues of variation operators that take into account the underlying graph connectivity. Smoothly varying signals defined on the nodes are of particular interest in various applications, and tend to be approximately bandlimited in the frequency basis. Sampling theory for graph signals deals with the problem of choosing the best subset of nodes for reconstructing a bandlimited signal from its samples. Most approaches to this problem require a computation of the frequency basis (i.e., the eigenvectors of the variation operator), followed by a search procedure using the basis elements. This can be impractical, in terms of storage and time complexity, for real datasets involving very large graphs. We circumvent this issue in our formulation by introducing quantities called graph spectral proxies, defined using the powers of the variation operator, in order to approximate the spectral content of graph signals. This allows us to formulate a direct sampling set selection approach that does not require the computation and storage of the basis elements. We show that our approach also provides stable reconstruction when the samples are noisy or when the original signal is only approximately bandlimited. Furthermore, the proposed approach is valid for any choice of the variation operator, thereby covering a wide range of graphs and applications. We demonstrate its effectiveness through various numerical experiments.
351 citations
Cites methods from "Compact Support Biorthogonal Wavele..."
...For example, spectral filters in the form of polynomials of the variation operator are used in the design of wavelet filterbanks for graph signals [21], [17], [18] to offer a trade-off between frequency-domain and vertex-domain localization....
[...]
...It considers a spectral-domain criterion, using minimum singular values of submatrices of the graph Fourier transform matrix, to minimize the effect of sample noise in the worst case....
[...]
TL;DR: This work formulate graph signal recovery as an optimization problem, for which it provides a general solution through the alternating direction methods of multipliers through which it relates to signal inpainting, matrix completion, robust principal component analysis, and anomaly detection.
Abstract: We consider the problem of signal recovery on graphs. Graphs model data with complex structure as signals on a graph. Graph signal recovery recovers one or multiple smooth graph signals from noisy, corrupted, or incomplete measurements. We formulate graph signal recovery as an optimization problem, for which we provide a general solution through the alternating direction methods of multipliers. We show how signal inpainting, matrix completion, robust principal component analysis, and anomaly detection all relate to graph signal recovery and provide corresponding specific solutions and theoretical analysis. We validate the proposed methods on real-world recovery problems, including online blog classification, bridge condition identification, temperature estimation, recommender system for jokes, and expert opinion combination of online blog classification.
271 citations
References
More filters
TL;DR: In this article, a structural similarity index is proposed for image quality assessment based on the degradation of structural information, which can be applied to both subjective ratings and objective methods on a database of images compressed with JPEG and JPEG2000.
Abstract: Objective methods for assessing perceptual image quality traditionally attempted to quantify the visibility of errors (differences) between a distorted image and a reference image using a variety of known properties of the human visual system. Under the assumption that human visual perception is highly adapted for extracting structural information from a scene, we introduce an alternative complementary framework for quality assessment based on the degradation of structural information. As a specific example of this concept, we develop a structural similarity index and demonstrate its promise through a set of intuitive examples, as well as comparison to both subjective ratings and state-of-the-art objective methods on a database of images compressed with JPEG and JPEG2000. A MATLAB implementation of the proposed algorithm is available online at http://www.cns.nyu.edu//spl sim/lcv/ssim/.
40,609 citations
TL;DR: The field of signal processing on graphs merges algebraic and spectral graph theoretic concepts with computational harmonic analysis to process high-dimensional data on graphs as discussed by the authors, which are the analogs to the classical frequency domain and highlight the importance of incorporating the irregular structures of graph data domains when processing signals on graphs.
Abstract: In applications such as social, energy, transportation, sensor, and neuronal networks, high-dimensional data naturally reside on the vertices of weighted graphs. The emerging field of signal processing on graphs merges algebraic and spectral graph theoretic concepts with computational harmonic analysis to process such signals on graphs. In this tutorial overview, we outline the main challenges of the area, discuss different ways to define graph spectral domains, which are the analogs to the classical frequency domain, and highlight the importance of incorporating the irregular structures of graph data domains when processing signals on graphs. We then review methods to generalize fundamental operations such as filtering, translation, modulation, dilation, and downsampling to the graph setting and survey the localized, multiscale transforms that have been proposed to efficiently extract information from high-dimensional data on graphs. We conclude with a brief discussion of open issues and possible extensions.
3,475 citations
TL;DR: In this paper, it was shown that under fairly general conditions, exact reconstruction schemes with synthesis filters different from the analysis filters give rise to two dual Riesz bases of compactly supported wavelets.
Abstract: Orthonormal bases of compactly supported wavelet bases correspond to subband coding schemes with exact reconstruction in which the analysis and synthesis filters coincide. We show here that under fairly general conditions, exact reconstruction schemes with synthesis filters different from the analysis filters give rise: to two dual Riesz bases of compactly supported wavelets. We give necessary and sufficient conditions for biorthogonality of the corresponding scaling functions, and we present a sufficient condition for the decay of their Fourier transforms. We study the regularity of these biorthogonal bases. We provide several families of examples, all symmetric (corresponding to “linear phase” filters). In particular we can construct symmetric biorthogonal wavelet bases with arbitrarily high preassigned regularity; we also show how to construct symmetric biorthogonal wavelet bases “close” to a (nonsymmetric) orthonormal basis.
2,854 citations
Book•
01 Mar 1995TL;DR: Wavelets and Subband Coding offered a unified view of the exciting field of wavelets and their discrete-time cousins, filter banks, or subband coding and developed the theory in both continuous and discrete time.
Abstract: First published in 1995, Wavelets and Subband Coding offered a unified view of the exciting field of wavelets and their discrete-time cousins, filter banks, or subband coding. The book developed the theory in both continuous and discrete time, and presented important applications. During the past decade, it filled a useful need in explaining a new view of signal processing based on flexible time-frequency analysis and its applications. Since 2007, the authors now retain the copyright and allow open access to the book.
2,793 citations
TL;DR: A novel method for constructing wavelet transforms of functions defined on the vertices of an arbitrary finite weighted graph using the spectral decomposition of the discrete graph Laplacian L, based on defining scaling using the graph analogue of the Fourier domain.
1,681 citations
"Compact Support Biorthogonal Wavele..." refers background in this paper
...However, it can be shown [2], that filters with polynomial spectral kernels can be implemented iteratively with one-hop operations at each node, and do not require matrix diagonalization....
[...]
...There has been a significant recent interest in the extension of wavelet transforms to graph signals, including wavelets on unweighted graphs for analyzing computer network traffic [5], diffusion wavelets and diffusion wavelet packets [6], [7], the “top-down” wavelet construction of [8], graph dependent basis functions for sensor network graphs [9], lifting based wavelets on graphs [10]–[12], multiscale wavelets on balanced trees [13], spectral graph wavelets [2], and our recent work on graph wavelet filterbanks [3]....
[...]
...This trade-off is being studied for graph signals as well [2]–[4]....
[...]