DrFabrizioSmeraldifrom School of Electronic Engineering and Computer Science, Queen Mary University of London Invited To Give a Talk

Title of talk: POKer: a Partial Order Kernel for comparing strings with alternative substrings

Abstract:

String alignment is a fundamental task in computational biology. While a variety of algorithms have been developed to handle independent or position-dependent symbol substitutions, the problem of block-wise substitutions remains challenging. In this talk I will describe a Partial Order Kernel, named POKer, defined as an exponentially weighted sum of local alignment scores. POKer can be used for comparison and classification of strings containing alternative substrings of variable length. The kernel is defined over the product of two directed acyclic graphs, each representing a string with alternative substrings, and is computed efficiently using dynamic programming. I will present an evaluation of the performance of POKer with Support Vector Machines on a dataset of strings generated by detecting overlapping motifs in a set of simulated DNA sequences. Compared to a generalization of a state-of-the-art string kernel, POKer achieves a higher classification accuracy.

Title of talk: Ranklets: orientation selective, non-parametric features for the detection of intensity and variance modulations in images

Abstract:

Rank features have long been known to the pattern recognition community as a robust tool suitable for conditions of high noise, low resolution or extreme variability. The popularity of rank features has however been limited by their reputation for being relatively coarse. Arguably one of the main drawbacks of most existing descriptors is the lack of orientation selectivity.

In this talk I introduce a family of rank features named Ranklets, that provide an orientation selective, multiscale representation of the image similar to that obtained with Haar wavelets. Ranklets are defined starting from the Mann-Whitney statistics and admit an intuitive combinatorial interpretation in terms of pairwise comparison of pixel values. The algorithm can easily be extended to hexagonal pixel lattices for application in embedded systems such as some digital cameras. A recently developed extension of Ranklets based on the Siegel-Tukey statistics allows detection of second-order stimuli; this has been shown to be useful for the processing of visual textures.

Ranklets have been successfully applied to point tracking, face detection, the processing of mammographic images and texture classification among others; I will discuss some of these applications in my talk.

DrFabrizioSmeraldi’s Biography:

Member of International Program Committee, SICE 2005 SICE 2005 is the International Conference on Instrumentation, Control and Information Technology to be held in August 8-10, 2005, Okayama University, Okayama, JAPAN .Foreign member, ASTA2 project The ASTA2 project on Statistical Learning: Theory, Algorithms and Applications is funded by the Italian Ministry of University and Research (2002-2005). Partners include eight Italian Universities/ Departments and three Foreign Members. Program co-chair, AVBPA 2001 The Third International Conference on Audio- and Video-based Biometric Person Authentication took place in Halmstad, Sweden, in June 2001 .Guest Editor, PRL Special Issue on Biometric Authentication Pattern Recognition Letters Vol 14, No 13, Sept 2003