Talks

• Simon Morgan. TV-L¹ and the flat norm for shape signatures and distance functions. CAAM, Rice University, 04/21/2008. [link]

Models and Algorithms for Inverse Problems

• A Curvilinear Search Method for p-Harmonic Flows on Spheres (with Dr. Donald Goldfarb and Zaiwen Wen) submitted. [CAAM TR08-02].


• Iteratively Reweighted Algorithms for Compressive Sensing (with Dr. Rick Chartrand) ICASSP 2008. [CAAM TR08-01]


• Bregman Iterative Algorithms for l₁-Minimization with Applications to Compressed Sensing (with Dr. Stanley Osher, Dr. Donald Goldfarb, and Dr. Jerome Darbon) SIAM J. Imaging Sciences 1(1), 143-168, 2008. [reprint]. [download the code]


• Fixed-Point Continuation for l₁-Minimization: Methodology and Convergence (with Dr. Elaine Hale and Dr. Yin Zhang) submitted. [pdf] (revised, Jan 2008). [download the code]


• An algorithm for l₁-Regularized Logistic Regression (with Jianing Shi and Dr. Stanley Osher) in preparation.


• Second-Order Cone Programming Methods for Total Variation Based Image Restoration (with Dr. Donald Goldfarb) SIAM J. Scientific Computing 27(2), 622-645, 2005. [reprint]. [download the code] (code requires Mosek)


• Slope and G-Set Characterization of Set-Valued Functions and Applications to Non-Differentiable Optimization Problems (with Dr. Otmar Scherzer and Dr. Stanley Osher) Communications in Mathematical Sciences 3(4), 479-492, 2005. [reprint]

Total Variation-Based Regularization for Image Processing

• Parametric Maximum Flow Algorithms for Fast Total Variation Minimization (with Dr. Donald Goldfarb) submitted. [CAAM TR07-09]. [download the code]


• A Fast Algorithm for Image Deblurring with Total Variation Regularization (with Yilun Wang and Dr. Yin Zhang) submitted. [CAAM TR07-10]. [download the code]


• A Comparison of Three Total Variation Based Texture Extraction Models (with Dr. Donald Goldfarb and Dr. Stanley Osher) Journal of Visual Communication and Image Representation, 18(3), 240-252, 2007. [reprint]


• The Total Variation Regularized L¹ Model for Multiscale Decomposition (with Dr. Donald Goldfarb and Dr. Stanley Osher) SIAM J. Multiscale Modeling and Simulation 6(1), 190-211, 2006. [reprint]


• Image Cartoon-Texture Decomposition and Feature Selection using The Total Variation Regularized L¹ Functional (with Dr. Donald Goldfarb and Dr. Stanley Osher) In Variational, Geometric, and Level Set Methods in Computer Vision, Lecture Notes in Computer Science 3752, 73-84, Springer, 2005. [reprint]

• An Iterative Regularization Method for Total Varitaion-Based Image Restoration (with Dr. Stanley Osher, Dr. Martin Burger, Dr. Donald Goldfarb, and Jinjun Xu) SIAM J. Multiscale Modeling and Simulation 4(2), 460-489, 2005. [reprint]

Medical Imaging and Computer Vision

• Background Correction for cDNA Microarray Images using The TV+L¹ Model (with Terrence Chen, Dr. Xiang Zhou, and Dr. Amit Chakraborty) Bioinformatics 21 (10), 2410-2416, 2005. [reprint]

• (Conf.) Illumination Normalization for Face Recognition and Uneven Background Correction using Total Variation Based Image Models (with Terrence Chen, Dr. Xiang Zhou, Dr. Dorin Comaniciu, and Dr. Thomas Huang) In proceedings of IEEE confererence on Computer Vision and Pattern Recognition (CVPR) 2005(2), 532-539. [reprint]

• Total Variation Models for Variable Lighting Face Recognition (with Terrence Chen, Dr. Xiang Zhou, Dr. Dorin Comaniciu, and Dr. Thomas Huang) IEEE Transactions on Pattern Analysis and Machine Intelligence (PAMI) 28 (9), 1519-1524, 2006. [reprint]

• A New Coarse-To-Fine Framework for 3D Brain MR Image Registration (with Terrence Chen, Dr. Thomas Huang, and Dr. Xiang Zhou) Computer Vision for Biomedical Image, Lecture Notes in Computer Science 3765, 114-124, Springer, 2005. [reprint]

Codes for Image Processing

• A SUPER FAST parametric maximum-flow binary code for MATLAB
  Code webpage: [link]
  
  Version 2.3, last update on Aug 08, 2007.
  More platforms and 3D images are now supported.
  Author: W. Yin
  


• A second-order cone programming code for MATLAB [link]
  
  Version 1, uploaded on 05/23/2006.
  Authors: D. Goldfarb and W. Yin @ Columbia University
  Require Mosek and Matlab preinstalled.