CCPR 2008 Keynote Speech 1

Summary form only given. Sensory information acquired by pattern recognition systems is invariably subject to environmental and sensing conditions, which may change over time. For imaging sensors, for instance, this includes illumination changes, pose and view-point changes, noise, distortion, blurring, etc. This has inevitably a significant negative impact on the performance of pattern recognition algorithms. In the past, these problems have been tackled one by one by the incorporation of ever increasing degree of invariance to such degradation phenomena in the representation of the sensory data. More recently, the possibility of enhancing pattern recognition system robustness by using auxiliary information has been explored. In particular, by measuring the extent of degradation, the resulting sensory data "quality" information can be used with advantage to combat the effect of degradation phenomena. This can be achieved either by using the auxiliary quality information as meta knowledge to control the sensory data interpretation process, or even as features in an augmented data representation space. Various approaches to, and the issues in quality based sensory data interpretation will be discussed. The problems and benefits associated with the use of auxiliary information in sensory data analysis will be illustrated on the problem of personal identity verification and recognition in biometrics, with a focus on multimodal biometrics approaches.