Sound examples of the paper are accessible with the following link:

Low bitrate informed source separation of realistic mixtures

by A. Liutkus, R.Badeau and G. Richard.

Demonstration is located at the following URL:

http://www.enst.fr/~liutkus/demoISS

Date: Mardi 27 novembre

hour: 11h

Room: Vitrine de la recherche (DA-006, Télécom ParisTech, site Dareau)

Abstract: Algorithms for decompositions of matrices are of central importance in machine learning, signal processing and information retrieval, with SVD and NMF (Nonnegative Matrix Factorisation) being the most widely used examples. Probabilistic interpretations of matrix [...]]]> Author: Taylan Cemgil

Date: Mardi 27 novembre

hour: 11h

Room: Vitrine de la recherche (DA-006, Télécom ParisTech, site Dareau)

Abstract:
Algorithms for decompositions of matrices are of central importance in machine learning, signal processing and information retrieval, with SVD and NMF (Nonnegative Matrix Factorisation) being the most widely used examples. Probabilistic interpretations of matrix factorisation models are also well known and are useful in many applications (Salakhutdinov and Mnih 2008; Cemgil 2009; Fevotte et. al. 2009). In the recent years, decompositions of multiway arrays, known as tensor factorisations have gained significant popularity for the analysis of large data sets with more than two entities (Kolda and Bader, 2009; Cichocki et. al. 2008). In particular, tensor models have been used in several
several audio processing applications. We will discuss a subset of these models from a statistical modelling perspective, building upon probabilistic generative models and generalised linear models (McCulloch and Nelder). In both views, the factorisation is implicit in a well-defined statistical model and factorisations can be computed via maximum likelihood.

We express a tensor factorisation model using a factor graph and the factor tensors are optimised iteratively. In each iteration, the update equation can be implemented by a message passing algorithm, reminiscent to variable elimination in a discrete graphical model. This setting provides a structured and efficient approach that enables very easy development of application specific custom models, as well as algorithms for the so called coupled (collective) factorisations where an arbitrary set of tensors are factorised simultaneously with shared factors. Extensions to full Bayesian inference for model selection, via variational approximations or MCMC are also feasible. Well known models of multiway analysis such as Nonnegative Matrix Factorisation (NMF), Parafac, Tucker, and audio processing (Convolutive NMF, NMF2D, SF-SSNTF) appear as special cases and new extensions can easily be developed. We will illustrate the approach with applications in audio and music processing.

Bio:
Taylan Cemgil received his Ph.D. (2004) from SNN, Radboud University Nijmegen, the Netherlands. Between 2004 and 2008 he worked as a postdoctoral researcher at the Signal Processing and Communications Lab., University of Cambridge, UK. He is currently an associate professor of Computer Engineering at Bogazici University, Istanbul, Turkey. He is interested in Bayesian statistical methods, machine learning and signal processing.

http://www.cmpe.boun.edu.tr/~cemgil/

Friday, september the 28th

Room: Vitrine de la recherche (DA-006, Télécom ParisTech, Dareau)

Hour: 12h45

On the Web page referenced below, you will have access to the source  code of  the  methods  proposed in the paper:

« Multiclass Feature Selection with Kernel Gram-matrix-based criteria  », M. Ramona, G. Richard, B. David, IEEE Transactions on Neural Networks and Learning Systems, to appear, 2012

We also  provide [...]]]> Multiclass Feature Selection with Kernel Gram-matrix-based criteria 

On the Web page referenced below, you will have access to the source  code of  the  methods  proposed in the paper:

« Multiclass Feature Selection with Kernel Gram-matrix-based criteria  », M. Ramona, G. Richard, B. David, IEEE Transactions on Neural Networks and Learning Systems, to appear, 2012

We also  provide further  details  on  the  features  of  the  speech/music  dataset used in this paper.

Link : Access to source codes

“A Flexible Multi-Resolution Time-Frequency Analysis Framework”, by Thomas Fillon and Jacques Prado, published in the proceedings of the 11th International Conference On Information Science, Signal Processing and their Applications (ISSPA) 3-5 july 2012, Montreal (Canada)

can be accessed at the following address [...]]]> The code corresponding to the time-frequency analysis framework described in the article :

“A Flexible Multi-Resolution Time-Frequency Analysis Framework”, by Thomas Fillon and Jacques Prado, published in the proceedings of the 11th International Conference On Information Science, Signal Processing and their Applications (ISSPA) 3-5 july 2012, Montreal (Canada)

can be accessed at the following address :

https://bitbucket.org/thomasfillon/mrt-toolbox

Contact : thomas<dot>fillon<at>telecom-paristech<dot>fr

This database has been gathered inside the QUAERO Project funded by the French agency OSEO.

This database is a joint work between the METISS team at INRIA/IRISA and the AAO team at Telecom-ParisTech.