VC's Strategic PhD Scholarship
Particle Swarm Optimisation for Image Recognition
Mark Johnston and Mengjie Zhang
Image recognition tasks arise in a wide variety of practical situations, e.g., detecting faces from video images, identifying suspected terrorists from fingerprint images and diagnosing medical conditions from X-rays. Particle Swarm Optimization (PSO) is a stochastic, population-based evolutionary algorithm for solving optimization and image classification problems. PSO uses ideas analogous to biological evolution and socialpsychological principles of behaviour to search the space of candidate solutions for a particular task. PSO has been applied to a variety of image recognition and optimisation tasks and has achieved a certain level of success. However, there are still limitations in particle topology and representation, search algorithms and feature discovery that restrict PSO for difficult image classification tasks.
Applicants are invited from excellent and enthusiastic students wishing to undertake a PhD in this area. A strong background in operations research, statistics, machine learning, computer science and C/C++/Java programming is desired.
The deadline for applications is 15 May, 2009. Application forms are available from www.victoria.ac.nz/scholarships.
Contact Dr Mark Johnston or A/Prof Mengjie Zhang for research proposals, and Philippa Hay for applications.