CALL FOR PARTICIPATION ADVANCED SCHOOL ON COMPUTER VISION, PATTERN RECOGNITION AND IMAGE PROCESSING (Organized by Vision, Image Processing and Sound Laboratory - Department of Computer Science - University of Verona) May 17-20,2004 Verona, Italy This school would like to be the first of a series of seminars, aiming at offering advanced lectures on significant topics related to Computer Vision, Pattern Recognition, and Image Processing. These schools are particularly addressed to PhD students, but all types of researchers are welcome. They will typically hold one week on one specific topic, so that audience can install a more productive interaction with the lecturer. The maximum number of participants is limited to 50 persons. In case of a larger number of applicants, priority will be given to PhD students. This first school focuses on the statistical approach in computer vision, analyzing the graphical models for learning and inference with related applications. Invited speaker: Prof. Brendan J. Frey, University of Toronto, Canada Course Title: Bayesian Networks and Algorithms for Inference and Learning: Applications in Computer Vision, Audio Processing, and Molecular Biology Description: Algorithms for automatically analyzing images, video, audio, communication signals, biological sequences, text, and other types of data should take into account the uncertain relationships between inputs, intermediate representations, and outputs. Probability theory can account for these uncertainties, and provides a way to pose information processing problems as the computational task of finding an appropriate probability model and computing conditional probabilities using the model. Complex probability models for real-world applications often involve millions of random variables and intractable density functions, so probabilities cannot be computed using straightforward approaches. This course examines the fundamental concepts of graph-based formulations of complex probability models and introduces computationally efficient techniques for computing probabilities and estimating parameters in these models. Although the course is a "fundamentals" course, we will study several impressive real-world applications. DAY 1 Bayesian networks, Markov random fields and factor graphs. Computing probabilities in graphical models, the elimination algorithm and the sum-product algorithm. Generative models. Case study: Computer Vision. DAY 2 Learning observed graphical models, the exponential family. Parameterized models, parameters as variables, models for classification, regression and clustering. Learning partially unobserved graphical models, free energy, iterative conditional modes, the EM algorithm. Case study: Molecular Biology. DAY 3 Mixtures of Gaussians, HMMs, the multivariate Gaussian, factor analysis, linear dynamic systems, Kalman filtering and smoothing, learning linear dynamic systems. Variational techniques, the sum-product algorithm in graphs with cycles, Bethe free energy. Case study: Audio Processing. DAY 4 Monte Carlo methods, rejection sampling, adaptive rejection sampling, importance sampling, particle filters, Markov chain Monte Carlo methods, Gibbs sampling, Metropolis algorithm. Wrap-up. Check out the following website for full details: http://vips.sci.univr.it/html/VIPSschool2004.html Please also distribute this to your PhD students/researchers/colleagues who might be interested. If there are any queries please email the following address: mailto:school_info@zeus.sci.univr.it >>>> Registration deadline is 30 April 2004 <<<<< Directors: Prof. Vittorio Murino, Dr. Andrea Fusiello Local Organizers: Andrea Colombari, Marco Cristani, Michela Farenzena