2009 IEEE International Workshop on Machine Learning for Signal Processing (MLSP 2009)

Editor(s): Tülay Adali, Jocelyn Chanussot, Christian Jutten, Jan Larsen

AbstractThe 2009 IEEE International Workshop on Machine Learning for Signal Processing (MLSP-2009) was held in Grenoble, France, September 2-4, 2009. Grenoble, capital of French Alps, and the neighboring towns represent about 580,000 inhabitants. The dynamism of the city is largely due to its position as the host of a large university, numerous higher schools of engineering and architecture with more than 60,000 students among which at least 15% are foreigners. With over than 15,000 researchers, Grenoble is the second main scientific center in France, including both public and large private research centers, the center for Atomic Research, and international facilities like the European synchrotron. In addition, Grenoble has a very unique geographical position, it is a surprisingly flat city, surrounded by three mountain chains, Chartreuse on the north, Belledonne on the east and Vercors on the south. We welcome everyone to Grenoble, and hope you enjoy your stay in the town, the beautiful surrounding area, and of course the conference.
The workshop is the nineteenth of a series of workshops sponsored by the IEEE Signal Processing Society and organized by the Machine Learning for Signal Processing Technical Committee (MLSP-TC). The first event in the series was the Neural Networks for Signal Processing Workshop (NNSP), which took place in Princeton, NJ, in 1991. Since September 2003, the name of the Technical Committee, and hence the title of the Workshop, is changed from Neural Networks for Signal Processing to Machine Learning for Signal Processing to better reflect the areas represented by the TC.
Machine learning is devoted to the design of algorithms able to learn from empirical data. This approach is especially important in signal and image processing, where sets of sensors, usually large and heterogeneous, provide large amounts of data, usually noisy. From a methodological point of view, machine learning is concerned with multi-dimensional and statistical signal processing, especially with problems like detection, estimation and optimization. In addition to classical supervised or unsupervised learning, reinforcement learning and semi-supervised learning, machine learning methods include Bayesian modeling, Markov models, support vector machines and kernel methods. From the applications point of view, it has a wide range of topics: adaptive filtering, pattern recognition, scene analysis in computer vision, data mining, robot control, data fusion, blind and semi-blind source separation, sparse component analysis, brain-computer interfaces, hyperspectral images, cognitive radio, etc.
At this year’s workshop, there were a total of 109 submitted papers, where one is a keynote paper and 22 are papers from authors experts in their areas who have been invited to write a contribution for special sessions. Among the 86 regular submitted papers, 56 have been accepted, corresponding to a 64 % acceptance rate, based on reports of at least three independent reviewers for each paper. We would like to thank the members of the program committee for their detailed reviews, which made it possible for us to make a careful selection and provide feedback to the authors, which helped improve the final versions of the papers that appear in the proceedings.
The program included theoretical contributions in signal detection, pattern recognition and classification, blind source separation, learning theory, Bayesian learning and modeling, and applied contributions in speech and audio processing, biomedical application and communications. The program also featured three Special Sessions, one on Brain-computer Interfaces organized by Prof. J. Millan, from Federal Polytechnic Institute from Lausanne, Switzerland, the second on Machine Learning in Remote Sensing Data Processing, organized by Prof. G. Camps-Valls, University of Valencia, Spain, and the third one on Learning in Markov Models organized by Prof. W. Pieczynski, Telecom SudParis, France and F. Forbes, INRIA Rhône-Alpes, France.
The program also included a tutorial talk on adaptive filtering for complex-valued data and application for fMRI, delivered by Prof. T. Adali, from University of Maryland, Baltimore County, USA, and Prof. V. D. Calhoun, from University of New Mexico, USA. Our warmest thanks go to our tutorial speakers, and to our plenary speakers, Prof. L. Hansen from Technical University of Denmark, Prof. K.-R. Müller from Technical University Berlin, Germany and Prof. J. Hérault, Université Joseph Fourier, Grenoble, France. We have also taken advantage of MLSP 2009 to honor Prof. J. Hérault, now emeritus professor in Grenoble University, for his pioneering contributions to machine learning, including providing the basis of blind source separation, high dimensional data analysis, and perception modeling. A special session, organized by Prof. A. Guérin-Dugué and funded by the French association Neurosciences and Engineering Sciences (NSI), with invited talks from Prof. M. Cottrell (Paris 1) and Dr. M. Verleysen (Mouvain-la Neuve, Belgium) has been organized for honoring Prof. J. Hérault.
Keywordsmachine learning, signal processing
TypeBook [Editorship]
Year2009    Month September
PublisherIEEE
ISBN / ISSNISBN 9781424449477
NoteDOI 10.1109/MLSP.2009.5306262
BibTeX data [bibtex]
IMM Group(s)Intelligent Signal Processing


Back  ::  IMM Publications