Grant: Strategic Projects - Group

The challenges in prevention of disasters, especially those associated with terrorism acts are immense and extremely timely. As of today, we witness a wave of rapidly developing highly sophisticated and diversified threats. Fortunately, there is a potential of exploiting a wealth of methodologies and relevant technologies that could bring effective means of counteracting those malicious activities. They support expectations with this regard so that the abilities of exploiting mechanisms of information technology to alleviate the problems in an efficient and effective manner are put in place. Sound mechanisms of recognition of individuals or groups of individuals are crucial to the prudent confinement of potential threats. Within the area of biometrics, face recognition techniques have occupied a highly visible role owing to a balanced approach when tacking several delicate matters including a formation of a sound tradeoff between high effectiveness (an ability to recognize individuals) and compliance with privacy issues. The role of innovative information technologies in biometrics, especially those emerging under the banner of Computational Intelligence (CI) and pattern recognition has been promising and these technologies have already demonstrated their potential. What makes the development of biometric systems and face recognition classification even more demanding are the rapidly growing databases of faces which substantially challenge existing classification schemes and calls for radical changes of the design methodology and further augmentation of the existing algorithmic framework. If the above challenges can be adequately addressed, it is evident that the construction of user-centric intelligent systems could bring a lot of tangible benefits. In close collaboration with our industrial partner, Global Biometrics which has an outstanding track record in the area of face recognition we will be bringing an advanced research methodology of face recognition and their related algorithmic developments. This proposal spans two key research topics within the Safety and Security target area: Event Detection and Risk and Vulnerability. The ultimate objective of this project is to develop a comprehensive methodology of face classification and produce a comprehensive algorithmic suite by taking advantage of the technology of Computational Intelligence and pattern recognition. The resulting prototype of a classification system will be extremely helpful in drawing sound conclusions of immediate practical relevance as to the suitability of some typical methods and their augmentation in presence of huge collections of faces involving thousands of individuals. Such a challenge, even though of significant practical relevance, has never been fully explored in the past.