A Numerical Tour of Signal Processing

A Numerical Tour of Signal Processing The Numerical Tours of Signal Processing, by Gabriel Peyré, gather Matlab, Python and Scilab experiments to explore modern signal and image processing. www.numerical-tours.com/ Tue, 20 Jun 2017 08:30:46 +0000 Tue, 20 Jun 2017 08:30:46 +0000 Jekyll v3.4.3 20 New Python Tours <p>A set of 20 new <a href="/python/">Numerical Tours</a> in Python has just been released. It consists in all the most important Numerical Tours, and covers all the topics, from Wavelet image denoising to 3D mesh parameterization. <a href="/python/">Enjoy!</a></p> <p>This conversion has been funded by the ERC project <a href="http://gpeyre.github.io/sigma-vision/">SIGMA-Vision</a>, and it has been performed by <a href="https://github.com/pierrestock">Pierre Stock</a>, congratulation for this nice work!</p> Sat, 17 Sep 2016 12:00:00 +0000 www.numerical-tours.com/2016/09/17/python-tours/ www.numerical-tours.com/2016/09/17/python-tours/ New Python Tours <p><a href="https://www.gipsa-lab.grenoble-inp.fr/~laurent.condat/">Laurent Condat</a> just released a very nice set of <a href="http://www.numerical-tours.com/python/">Python tours</a> on optimization and applications in imaging. Simply go to the <a href="http://www.numerical-tours.com/python/">“optimization” section</a> in the Python’s tours. These new tours cover a wide range of topics, but target in particuler non-smooth convex optimization and first order proximal splitting schemes. Thanks <a href="https://www.gipsa-lab.grenoble-inp.fr/~laurent.condat/">Laurent</a>!</p> Tue, 14 Jun 2016 12:00:00 +0000 www.numerical-tours.com/2016/06/14/python-optimization/ www.numerical-tours.com/2016/06/14/python-optimization/ Updated tour <p>I have updated the tour on <a href="/matlab/optimaltransp_5_entropic/">Entropic Regularization of Optimal Transport</a>. It now present the algorihtms the way they should be presented, i.e. using the concept of diagonal scaling, that corresponds to Sinkhorn’s algorithm in the case of transport, and that can be extended to the computation of barycenters.</p> Mon, 21 Sep 2015 12:00:00 +0000 www.numerical-tours.com/2015/09/21/updated-tour-entropic/ www.numerical-tours.com/2015/09/21/updated-tour-entropic/ Kernel Update <p><a href="https://github.com/blink1073">Steven Silvester</a> converted to the whole set of tours to the <a href="https://jupyter.org/">Jupyter notebooks</a>. This means in particular that Matlab tours now run natively with the use of the <a href="https://github.com/calysto/matlab_kernel">Matlab kernel</a>.</p> Thu, 28 May 2015 12:00:00 +0000 www.numerical-tours.com/2015/05/28/jupyter-kernel/ www.numerical-tours.com/2015/05/28/jupyter-kernel/ New tour <p><a href="/matlab/meshproc_7_geodesic_poisson">This tour details</a> the <a href="http://dl.acm.org/citation.cfm?id=2516977">Geodesic in Heat method</a>, which allows one to approximate geodesic distances by solving a Poisson equation. It relies on the powerful result that the level sets of the heat diffusion converge to those of the geodesic distance as the diffusion time tends to 0.</p> Sun, 23 Nov 2014 12:00:00 +0000 www.numerical-tours.com/2014/11/23/new-tour-geodesic-poisson/ www.numerical-tours.com/2014/11/23/new-tour-geodesic-poisson/ Julia notebooks <p>The <a href="/julia/">first Numerical Tour</a> in <a href="http://julialang.org/">Julia</a> is available. <a href="http://julialang.org/">Julia</a> is an amazing language for scientific computing, that combines the ease of use of Matlab with speed comparable of C and an ecosystem which includes the powerful IPython notebooks.</p> Sun, 09 Nov 2014 12:00:00 +0000 www.numerical-tours.com/2014/11/09/julia-notebooks/ www.numerical-tours.com/2014/11/09/julia-notebooks/ Matlab notebooks <p>Thanks to <a href="https://github.com/blink1073">Steven Silvester</a>, most of the <a href="/matlab/">Matlab tours</a> are now available as iPython notebooks. This means that:</p> <ul> <li>The corresponding tours are now directly rendered online using <a href="http://nbviewer.ipython.org/">nbviewer</a>, which means that HTML conversion is not anymore needed.</li> <li>The user can now download the corresponding .ipynb file and run it locally, possibly modifying and completing its content.</li> </ul> <p>This requires that you install Python and iPython (I recommend for instance the <a href="http://continuum.io/downloads">Anaconda</a> distribution). You also need to install the <a href="https://pypi.python.org/pypi/pymatbridge">pymatbridge</a> Python module in order to run Matlab code from the notebook.</p> <p>The command</p> <blockquote> <p>%load_ext pymatbridge</p> </blockquote> <p>will start Matlab in the background, and then you simply need to put the keyword</p> <blockquote> <p>%%matlab</p> </blockquote> <p>if you want to add a new cell to the notebook with your own Matlab code.</p> Sat, 01 Nov 2014 12:00:00 +0000 www.numerical-tours.com/2014/11/01/matlab-notebooks/ www.numerical-tours.com/2014/11/01/matlab-notebooks/ New site <p>A lifting for the web site, which is now hosted on the github project pages.</p> Mon, 20 Oct 2014 12:00:00 +0000 www.numerical-tours.com/2014/10/20/new-site/ www.numerical-tours.com/2014/10/20/new-site/ New tour <p><a href="/matlab/sparsity_9b_music/">This tour explores</a> the use of the MUSIC algortithm to perform sparse spikes deconvolution. The localization of the spikes is obtained by finding the root of a polynomial.</p> Tue, 26 Aug 2014 12:00:00 +0000 www.numerical-tours.com/2014/08/26/new-tour-music/ www.numerical-tours.com/2014/08/26/new-tour-music/ New tour <p><a href="/matlab/optimaltransp_5_entropic/">This tour explores</a> a fast algorithm to compute the solution of a regularized optimal transportation problem. This corresponds to the resolution of a convex program which is the projection on the transport polytope for a Kulback-Leibler divergence.</p> Tue, 26 Aug 2014 12:00:00 +0000 www.numerical-tours.com/2014/08/26/new-tour-entropic/ www.numerical-tours.com/2014/08/26/new-tour-entropic/