Contact details

Research

Cospectrality and graph isomorphism: One way of distinguishing between graphs is to use the spectrum of a matrix representation of the graph. We have done a number of studies looking at the properties of representations and their spectra. Some are better than others... 

Spectral Methods for Graphs: We have recently shown how eigenvectors of the graph Laplacian can be used to construct features which describe a graph. However, this feature space is complex and needs sophisticated tools to analyse it. This work is aimed at examining the properties of these features and applying them to pattern recognition problems with graphs.

Chemical Structure Databases: Databases of chemical structures are an important tool for the pharmaceutical industry because they allow virtual screening of molecules for particular biological effects. The molecules in these databases are typically represented by atom-bond graphs. The aim of this work is to apply graph matching methods to retrieval, activity  prediction and generation of new molecules.

Stereo and Shape from Shading: Stereo algorithms can be used to extract 3D structure from a pair (or more) of images. Stereo is very effective where good correspondences can be found, i.e. where there is clear surface texture, but fails on smooth surfaces. Shape from shading can also be used to extract 3D information from a single image, but only for textureless surface. The aim of this work is to combine these two methods to produce a more effective and complete reconstruction.

Relational Graph Matching: This work is aimed at matching graphs by providing principled Bayesian methods for rectifying initialisation errors, in the presence of significant structural error in the graphs. The second part of this research provides techniques for identifying and removing structural errors from relational graphs.

Current projects

SIMBAD: EU project.

Publications/CV

You can access and search the group publications via the group homepage. Some of these publications are available for download.

Teaching

The CVI module (third year option) covers topics in computer vision. The course pages are here.

The PAT module (third year option) is a course on statistical pattern recognition. The course pages are here.

The PRG module (MEng module) is a group programming project. The course pages are here.

My undergraduate projects for 2008.