Hong-Yuan Mark Liao, Chin-Chuan Han and Gwo-Jong Yu
Face recognition, by definition, should be a recognition process in which recognition is based on the content of a face. The problem is: what is a ``face'' ? Goudail et al. and Swets and Weng have recently proposed state-of-the-art face recognition systems. Through careful implementation, the results have shown that both methods are valid and efficient. However, we have a question about the ``face'' images they have adopted. They used ``face'' images that included hair, shoulders, face and background. Our intuition tells us that only a recognition process based on a ``pure'' face portion can be called face recognition. From a series of experiments, we have shown that the ``real'' face portion in their face recognition process did not play a key role at all. Instead, the combination of hair, shoulder and background dominated the whole recognition process. We suggest that future research on face recognition should use a face-only database, not a face combined with other irrelevant portions.