Novel face recognition based on individual eigen-subspaces

In this paper, a novel method of face recognition based on individual eigen-subspaces is presented, which is expected to tackle such problems on face recognition as the ready availability of reference samples for each subject. In the proposed method, multiple face eigensubspaces are created, with each one corresponding to one known subject privately, rather than all individuals sharing one universal subspace as in the traditional eigenface method. Compared with the traditional single subspace face representation, the proposed method captures the extrapersonal difference to the most possible extent, which is crucial to distinguish between individuals, and on the other hand, it throws away the most intrapersonal difference and noise in the input. Our experiments strongly support the proposed idea, in which 20% improvement of performance over the traditional "eigenface" has been observed when tested on the same face base.