In this paper, in order to achieve high-speed PCA, the simple principal component analysis (SPCA) is applied to compress the dimensionality of portions that constitute a face. By using neural networks, the difference in the value of cos θ between the true and the false smiles is clarified and the true smile is discriminated. Finally, in order to show the effectiveness of the proposed face classification method for true or false smiles, computer simulations are done with real images. Furthermore, an experiment using the self-organization map is also conducted as a comparison.
共著者：Miyoko Nakano, Fumiko Yasukata, Minoru Fukumi
