Finalist for the SET Awards
September 2009
Timothy Yorke, one of our First Class Honours graduates and a project prize recipient, has been shortlisted for the Best Electronic Engineering Student Awards 2009 with his third year project entitled “Facial Recognition Using Neural Networks”
The SET Awards (Science, Engineering & Technology Student of the Year) are Britain's most important awards for science and technology undergraduates. The SET awards provide a showcase for educational excellence by publicly recognising the exceptional achievements of both students and universities. Each category selects three finalists from submissions around the country. The finalists will be interviewed and the winners will be announced at the Presentation Ceremony, to be held in London on the 24th September 2009. More information can be found on the SET Awards website (http://www.setawards.org)
In submitting Tim for the SET Awards, Dr Hujun Yin, his project supervisor, wrote “Tim always came to the weekly regular meeting with good progress to report and targets for the next, and managed the project well. He is highly motivated, well-organised, hard working, and most importantly he always delivers! It was a pleasure to have him as a project student. In my 12 years of lecturing and supervision at Manchester, I have had many project students achieving first class project work and first class degree. However, Tim is head and shoulders above many first class students. The amount of the work and quality of the work are by far the best I have seen.”
Tim started at this School in September 2006 on a 3-year Honours Course (BEng in Electrical and Electronic Engineering). His performance in the third year was the highlight of his ability and achievement here at Manchester, with the average mark of all exam modules over 80%! His overall project mark (assessed on six deliverables over two semesters by two academics) was 88% - one of the highest in the School. Tim chose this challenging project as he thought was a good realisation of electronic engineering and information technology. He started to show his skills and ability towards the end of the first semester. He had not only mastered all the fundamental components of the projects from dimensionality reduction, image pre-processing, face detection and extraction, to pattern classification, but also implemented them. In the second semester, he made tremendously progress – well exceeding the expectation. He had researched into various neural networks and their advantages and disadvantages and implemented and tested them on benchmark datasets. He accomplished all the objectives set out in the project description at the beginning of the project and he went further to elevate the project to a higher level and standard. He added some novel elements into the project such as online reconfiguration and training and efficient classification procedure. He further studied the effect of lighting and pre-processing. He built a real-time demo system that demonstrates the fully capability of the neural based adaptive system on a small-medium set of locally sampled subjects. The developed software and implemented components had been seamlessly integrated with the hardware and camera. The system is practical and works in real-time recognition.
The project was part of a long-term, on-going research theme in Dr. Yin’s group. The research has recently attracted the attention of UMIP – University of Manchester Intellectual Property Commercialisation Company, who are currently evaluating the potential of further development for security markets.