E&CSE Research Seminar, Thursday 15 - November - 2007

Abstract:

With the advent and development of new sensing technologies, it is now possible to capture image data in tens or hundreds of bands covering a broad spectral range. Spectral imaging provides an information-rich representation of the object under study and has a wide variety of applications in biosecurity, ecology, human performance, health, food quality assurance and surveillance. In this talk, I will elaborate on the development of robust spectroscopy techniques, large-scale databases, robust feature extraction algorithms and reliable classifiers for recognition and identification. I will explore the use of spectral absorption features for purposes of identification and recognition and show how discriminant learning can be used to perform a localised analysis of the spectra. Under this framework,  multiclass classification problems can be addressed using a pairwise discriminant analysis algorithm based upon a novel cost function. If time permits, I will also elaborate on how the optimisation problem can be solved making use of unconstrained techniques over Grassmann manifolds. This approach does not require matrix inversion operations and, therefore, does not suffer of stability problems for small training sets.

About the Speaker:

Antonio Robles-Kelly received his B.Eng. degree in Electronics and Telecommunications from the Inst. Tecnologico y de Estudios Superiores de Monterrey with honours in 1998. In 2001, being a graduate student at York , he visited the University of South Florida as part of the William Gibbs/Plessey Award to the best research proposal to visit an overseas research lab. He received his PhD in Computer Science from the University of York in 2003. After receiving his doctorate, Dr. Robles-Kelly remained in York until Dec. 2004 as a Research Associate under the MathFit-EPSRC framework. In 2005, he took a research scientist appointment with National ICT Australia (NICTA) at the Canberra   Lab, where he leads the Spectral Imaging and Source Mapping (SISM) project. He is also a Postdoctoral Research Fellow of the Australian Research Council and an Adjunct Research Fellow at the ANU. His research interests are in the areas of Computer Vision, Pattern Recognition, Spectral Imaging and Computer Graphics. Along these lines, he has done work on segmentation and grouping, tracking, graph-matching, shape-from-X, hyperspectral image understanding, reflectance and graphical models for recognition and classification.