Rank: 688 based on 166 downloads (last 30 days) and 4 files submitted
photo

Dominique Zosso

E-mail
Company/University
UCLA Mathematics

Personal Profile:

http://www.math.ucla.edu/~zosso

Professional Interests:
Inverse Problems in Image Processing, Computer Vision, Machine Learning

 

Watch this Author's files

 

Files Posted by Dominique View all
Updated   File Tags Downloads
(last 30 days)
Comments Rating
18 Aug 2014 Screenshot Non-local retinex MATLAB implementation of generalized non-local Retinex (contrast enhancement, shadow removal, etc.) Author: Dominique Zosso retinex, nonlocal, shadow removal, shadow detection, contrast enhancement, dynamic range compres... 45 1
06 Aug 2014 Screenshot Efficient Beltrami Image Denoising and Deconvolution Implementation of our Primal-Dual Projected Gradients algorithm for efficient Beltrami regularizatio Author: Dominique Zosso beltrami, image denoising, nonblind deconvolutio..., deconvolution, image restoration, regularization 35 0
18 Mar 2014 Screenshot Two-dimensional Variational Mode Decomposition Variationally decompose a 2D signal into k band-separated modes. Author: Dominique Zosso adaptive filter, band separated, bandwidth, center frequency, decomposition, emd 35 1
20 Dec 2013 Screenshot Variational Mode Decomposition Variationally decompose a 1D signal into k band-separated modes. Author: Dominique Zosso variational mode deco..., vmd, bandwidth, band separated, adaptive filter, decomposition 51 2
  • 5.0
5.0 | 1 rating
Comments and Ratings by Dominique View all
Updated File Comments Rating
15 Aug 2014 Non-local retinex MATLAB implementation of generalized non-local Retinex (contrast enhancement, shadow removal, etc.) Author: Dominique Zosso

Check http://www.math.ucla.edu/~zosso for updates on the retinex publications

22 Jul 2014 Variational Mode Decomposition Variationally decompose a 1D signal into k band-separated modes. Author: Dominique Zosso

@Timothy Scharf:

there is no sharp rule for alpha, since the value essentially depends on the amplitude of signals and the noisiness of the data. Trial and error, mostly, I'm afraid.

For the tau (time step): this governs how quickly the Lagrangian multiplier accumulates the "reconstruction error". We have mostly worked with values of either 0 (no accumulation, the reconstruction is not strictly enforced, but encouraged in least-squares sense, allowing for the noise to go in these residuals), or around 0.1. Higher values might lead to overly fast "freezing" of the modes.

Hope that helps.
Dominique

18 Mar 2014 Two-dimensional Variational Mode Decomposition Variationally decompose a 2D signal into k band-separated modes. Author: Dominique Zosso

Small glitch in the visualization part of the test file:

Where the evolution of the center frequencies is traced, the plot-line should read as follows:

plot( size(f,2)*(0.5+omega(:,1,k)), size(f,1)*(0.5+omega(:,2,k)), colors(k) );

(size(f,1) and size(f,2) need to be swapped).

This does not affect the decomposition itself, however.

24 Jul 2013 matlab2tikz A script to convert MATLAB/Octave into TikZ figures for easy and consistent inclusion into LaTeX. Author: Nico Schlömer

Absolutely great tool!

The whitespace issues (cf Christoph, March 25, 2013) have not entirely been fixed, though (v.0.4.0):

Around the "definecolor" lines, some more '%' should be placed:

line 555: put an extra %% in front

m2t.content.colors = sprintf('%%\n%% defining custom colors\n');

line 562, too:

m2t.content.colors = [m2t.content.colors sprintf('%%\n')];

Thanks for adding this in future updates! Great work!

Comments and Ratings on Dominique's Files View all
Updated File Comment by Comments Rating
15 Aug 2014 Non-local retinex MATLAB implementation of generalized non-local Retinex (contrast enhancement, shadow removal, etc.) Author: Dominique Zosso Zosso, Dominique

Check http://www.math.ucla.edu/~zosso for updates on the retinex publications

22 Jul 2014 Variational Mode Decomposition Variationally decompose a 1D signal into k band-separated modes. Author: Dominique Zosso Zosso, Dominique

@Timothy Scharf:

there is no sharp rule for alpha, since the value essentially depends on the amplitude of signals and the noisiness of the data. Trial and error, mostly, I'm afraid.

For the tau (time step): this governs how quickly the Lagrangian multiplier accumulates the "reconstruction error". We have mostly worked with values of either 0 (no accumulation, the reconstruction is not strictly enforced, but encouraged in least-squares sense, allowing for the noise to go in these residuals), or around 0.1. Higher values might lead to overly fast "freezing" of the modes.

Hope that helps.
Dominique

04 Jun 2014 Variational Mode Decomposition Variationally decompose a 1D signal into k band-separated modes. Author: Dominique Zosso Scharf, Timothy

Thank you for posting this.

I am tinkering around with your algorithm for EEG classification for seizure detection (in lieu of the very common EMD) .

I wonder if you could shed any light on the alpha (band width constraint) and tau (time-step) parameters? I have read and re-read the paper and still need a little guidance from the author. Thanks either way for posting

Tim

18 Mar 2014 Two-dimensional Variational Mode Decomposition Variationally decompose a 2D signal into k band-separated modes. Author: Dominique Zosso Zosso, Dominique

Small glitch in the visualization part of the test file:

Where the evolution of the center frequencies is traced, the plot-line should read as follows:

plot( size(f,2)*(0.5+omega(:,1,k)), size(f,1)*(0.5+omega(:,2,k)), colors(k) );

(size(f,1) and size(f,2) need to be swapped).

This does not affect the decomposition itself, however.

Contact us