From Sparse Solutions of Systems of Equations to Sparse Modeling of Signals and Images Alfred M. Bruckstein (Technion), David L. Donoho (Stanford), Michael.

Slides:



Advertisements
Similar presentations
Compressive Sensing IT530, Lecture Notes.
Advertisements

Visual Dictionaries George Papandreou CVPR 2014 Tutorial on BASIS
Slide 1 Bayesian Model Fusion: Large-Scale Performance Modeling of Analog and Mixed- Signal Circuits by Reusing Early-Stage Data Fa Wang*, Wangyang Zhang*,
Multi-Label Prediction via Compressed Sensing By Daniel Hsu, Sham M. Kakade, John Langford, Tong Zhang (NIPS 2009) Presented by: Lingbo Li ECE, Duke University.
MMSE Estimation for Sparse Representation Modeling
Joint work with Irad Yavneh
Pixel Recovery via Minimization in the Wavelet Domain Ivan W. Selesnick, Richard Van Slyke, and Onur G. Guleryuz *: Polytechnic University, Brooklyn, NY.
Learning Measurement Matrices for Redundant Dictionaries Richard Baraniuk Rice University Chinmay Hegde MIT Aswin Sankaranarayanan CMU.
Online Performance Guarantees for Sparse Recovery Raja Giryes ICASSP 2011 Volkan Cevher.
Submodular Dictionary Selection for Sparse Representation Volkan Cevher Laboratory for Information and Inference Systems - LIONS.
The Analysis (Co-)Sparse Model Origin, Definition, and Pursuit
Sparse Representations and the Basis Pursuit Algorithm* Michael Elad The Computer Science Department – Scientific Computing & Computational mathematics.
K-SVD Dictionary-Learning for Analysis Sparse Models
Wangmeng Zuo, Deyu Meng, Lei Zhang, Xiangchu Feng, David Zhang
Extensions of wavelets
* * Joint work with Michal Aharon Freddy Bruckstein Michael Elad
More MR Fingerprinting
Ilias Theodorakopoulos PhD Candidate
An Introduction to Sparse Coding, Sparse Sensing, and Optimization Speaker: Wei-Lun Chao Date: Nov. 23, 2011 DISP Lab, Graduate Institute of Communication.
Compressed sensing Carlos Becker, Guillaume Lemaître & Peter Rennert
Sparse & Redundant Signal Representation, and its Role in Image Processing Michael Elad The CS Department The Technion – Israel Institute of technology.
Dictionary-Learning for the Analysis Sparse Model Michael Elad The Computer Science Department The Technion – Israel Institute of technology Haifa 32000,
Sparse Representation and Compressed Sensing: Theory and Algorithms
Volkan Cevher, Marco F. Duarte, and Richard G. Baraniuk European Signal Processing Conference 2008.
Sparse and Overcomplete Data Representation
SRINKAGE FOR REDUNDANT REPRESENTATIONS ? Michael Elad The Computer Science Department The Technion – Israel Institute of technology Haifa 32000, Israel.
Mathematics and Image Analysis, MIA'06
Image Denoising via Learned Dictionaries and Sparse Representations
An Introduction to Sparse Representation and the K-SVD Algorithm
Optimized Projection Directions for Compressed Sensing Michael Elad The Computer Science Department The Technion – Israel Institute of technology Haifa.
Image Denoising with K-SVD Priyam Chatterjee EE 264 – Image Processing & Reconstruction Instructor : Prof. Peyman Milanfar Spring 2007.
Random Convolution in Compressive Sampling Michael Fleyer.
Introduction to Compressive Sensing
Analysis of the Basis Pustuit Algorithm and Applications*
Recent Trends in Signal Representations and Their Role in Image Processing Michael Elad The CS Department The Technion – Israel Institute of technology.
Image Decomposition and Inpainting By Sparse & Redundant Representations Michael Elad The Computer Science Department The Technion – Israel Institute of.
Compressive sensing: Theory, Algorithms and Applications
Over-Complete & Sparse Representations for Image Decomposition*
A Weighted Average of Sparse Several Representations is Better than the Sparsest One Alone Michael Elad The Computer Science Department The Technion –
A Sparse Solution of is Necessarily Unique !! Alfred M. Bruckstein, Michael Elad & Michael Zibulevsky The Computer Science Department The Technion – Israel.
Topics in MMSE Estimation for Sparse Approximation Michael Elad The Computer Science Department The Technion – Israel Institute of technology Haifa 32000,
Alfredo Nava-Tudela John J. Benedetto, advisor
Compressed Sensing Compressive Sampling
An ALPS’ view of Sparse Recovery Volkan Cevher Laboratory for Information and Inference Systems - LIONS
Compressive Sampling: A Brief Overview
AMSC 6631 Sparse Solutions of Linear Systems of Equations and Sparse Modeling of Signals and Images: Midyear Report Alfredo Nava-Tudela John J. Benedetto,
Game Theory Meets Compressed Sensing
Cs: compressed sensing
Introduction to Compressive Sensing
“A fast method for Underdetermined Sparse Component Analysis (SCA) based on Iterative Detection- Estimation (IDE)” Arash Ali-Amini 1 Massoud BABAIE-ZADEH.
Introduction to Compressed Sensing and its applications
Fast and incoherent dictionary learning algorithms with application to fMRI Authors: Vahid Abolghasemi Saideh Ferdowsi Saeid Sanei. Journal of Signal Processing.
Eran Treister and Irad Yavneh Computer Science, Technion (with thanks to Michael Elad)
Model-Based Compressive Sensing Presenter: Jason David Bonior ECE / CMR Tennessee Technological University November 5, 2010 Reading Group (Richard G. Baraniuk,
Learning to Sense Sparse Signals: Simultaneous Sensing Matrix and Sparsifying Dictionary Optimization Julio Martin Duarte-Carvajalino, and Guillermo Sapiro.
Sparse & Redundant Representation Modeling of Images Problem Solving Session 1: Greedy Pursuit Algorithms By: Matan Protter Sparse & Redundant Representation.
A Weighted Average of Sparse Representations is Better than the Sparsest One Alone Michael Elad and Irad Yavneh SIAM Conference on Imaging Science ’08.
Jianchao Yang, John Wright, Thomas Huang, Yi Ma CVPR 2008 Image Super-Resolution as Sparse Representation of Raw Image Patches.
Jeremy Watt and Aggelos Katsaggelos Northwestern University
Basic Algorithms Christina Gallner
Presented by: Mingyuan Zhou Duke University, ECE Feb 22, 2013
A Motivating Application: Sensor Array Signal Processing
Sparse Regression-based Hyperspectral Unmixing
Optimal sparse representations in general overcomplete bases
Sparse and Redundant Representations and Their Applications in
* * Joint work with Michal Aharon Freddy Bruckstein Michael Elad
CIS 700: “algorithms for Big Data”
LAB MEETING Speaker : Cheolsun Kim
Outline Sparse Reconstruction RIP Condition
Presentation transcript:

From Sparse Solutions of Systems of Equations to Sparse Modeling of Signals and Images Alfred M. Bruckstein (Technion), David L. Donoho (Stanford), Michael Elad (Technion) SIAM REVIEW 2009 Presented by: Mingyuan Zhou Duke University, ECE June 11, 2009

Outline Introduction The sparsest solution of Ax = b Variations on P 0 Sparsity-seeking methods in signal processing Processing of sparsely generated signals Applications in image processing

Under-determined linear system equation L2 norm L0 norm How can uniqueness of a solution be claimed? How to verify a candidate solution? How to efficiently solve the problem (the exhaustive search is a NP-hard problem)? What kind of approximations will work and how accurate can those be? Introduction

Conditions under which has a unique solution Conditions under which has the unique solution as Conditions under which the solution can be found by some “pursuit” algorithm Less restrictive notions of sparsity, impact of noise, the behavior of approximate solutions, and the properties of problem instances defined by ensembles of random matrices… Current achievements

JPEG, DCT JPEG-2000, DWT The sparsity of representation under given basis is key to many important signal and image processing problems: Image compression, Image denoising, image deblurring, speech compression, audio compression… The signal processing perspective Measuring sparsity

The sparsest solution of Ax = b

Uniqueness via the Spark Uniqueness via the Mutual Coherence Uniqueness

Greedy Algorithms Convex Relaxation Techniques Pursuit Algorithms: Practice

Greedy Algorithms Convex Relaxation Techniques Pursuit Algorithms: Performance

Uncertainty Principles and Sparsity Variations on P 0

From Exact to Approximate Solutions Relaxed constraint: Stability: Pursuit algorithms: OMP Iteratively reweighted least squares (IRLS) Iterative thresholding Stepwise algorithms: LARS and Homotopy

Performance of pursuit algorithms

Beyond Coherence Arguments Without noise With noise Empirical evidence: The column of A is drawn at random from a Gaussian distribution,,

Phase transitions in typical behavior:

Phase transitions in typical behavior:

Restricted isometry property (RIP):

The sparsest solution of Ax = b: A summary Uniqueness Solvability Approximate solutions Beyond coherence

Sparsity-Seeking Methods in Signal Processing Non-Gaussian Prior Combined representation

Processing of Sparsely Generated Signals Applications Analysis Compression Denoising Inverse problems Compressive sensing Morphological Component Analysis

The quest for a dictionary Reconstructed dictionaries Dictionaries learned from training data Dictionaries learned from data under test Learning Methods: MOD, K-SVD, BPFA

Applications in Image Processing Compression of Facial Images

Denoising of Images

Denoising of Images

Summary