Title :
A subspace method for feature extraction using independent components
Author :
Kinukawa, Shuhei ; Kotani, Manabu ; Ozawa, Seiichi
Author_Institution :
Fac. of Eng., Kobe Univ., Japan
Abstract :
We propose an approach to feature extraction using independent component analysis. The proposed method is combination of a subspace method and a combining strategy. Simulations are performed for the standard databases. Results reveal the effectiveness of the proposed method.
Keywords :
feature extraction; independent component analysis; pattern recognition; feature extraction; independent component analysis; independent components; primary visual cortex; recognition; subspace method; Data mining; Feature extraction; Humans; Image databases; Independent component analysis; Layout; Pattern recognition; Principal component analysis; Spatial databases; Vectors;
Conference_Titel :
SICE 2002. Proceedings of the 41st SICE Annual Conference
Print_ISBN :
0-7803-7631-5
DOI :
10.1109/SICE.2002.1195249
