In this paper, the simple-IFLDA of yielding features suitable for pattern recognition was presented by introducing further improvement into the simple-FLDA. The simple-IFLDA is derived from geometrical interpretation of the conventional FLDA as well as the simple-FLDA.
統計的学習アルゴリズムであるsimple-FLDAを用いて顔画像処理を行い,コンピュータシミュレーションにより有効性を検証した..
共同発表者:Miyoko Nakano, Stephen
Karungaru, Satoru Tsuge, Minoru Fukumi