Examinat ion of Unlearned Face in a Rotation and Size Spreading Discriminat ion Neural Network
Tsukasa Sakamoto, Hironobu Takano, Kiyomi Nakamura · 2004
Emulating the parietal cortex, a rotation and size spreading associative neural net- work (RS-SAN net) was developed. We extended the original system to make a human face recognition system which learned and recollected human face images. The recognition characteris- tics of the RS-SAN net were investigated for learned faces. However, we used the same human face images in the learning and recollection processes. In the present study, we investigated the recog- nition characteristics using different face images of the same subject. In addition, we investigated the recognition characteristics for unlearned faces. The original system was improved to make the recognition system cape with unlearned faces. We introduced a inner product and a minimum distance as shape recognition criteria. By setting the threshold ranges of the inner product and the minimum distance as 0.999 - 0.997 and 0.04 - 0.08, respectively, the false rejection and the false acceptance rate became 0% in both criteria.