Title :
Supervised independent component analysis by maximizing relative entropy
Author :
Yaping Huang ; Luo, Siwci ; Qi, Yingjian
Author_Institution :
Dept. of Comput. Sci. & Technol., Beijing Jiao Tong Univ., China
fDate :
31 Aug.-4 Sept. 2004
Abstract :
This paper presents a novel algorithm called SICA-MRE (supervised independent component analysis by maximizing relative entropy). It removes the disadvantage in traditional ICA, which ignores the contributions of independent components to recognition performance. Experimental results in face and iris recognition show that the presented algorithm has better performance.
Keywords :
entropy; eye; face recognition; independent component analysis; learning (artificial intelligence); face recognition; iris recognition; maximizing relative entropy; supervised independent component analysis; Biomedical computing; Biomedical signal processing; Blind source separation; Entropy; Face recognition; Independent component analysis; Iris; Signal analysis; Statistical analysis; Virtual colonoscopy;
Conference_Titel :
Signal Processing, 2004. Proceedings. ICSP '04. 2004 7th International Conference on
Print_ISBN :
0-7803-8406-7
DOI :
10.1109/ICOSP.2004.1441635
