id	author	title	date	pages	extension	mime	words	sentences	flesch	summary	cache	txt
work_igv6r4iavzgn3mtitipb2naxam	Xiaolong Fan	Selection and fusion of facial features for face recognition	2009	13	.pdf	application/pdf	9268	1099	73	The goal of the proposed technique is to select the most significant facial features effectively and find the best combination of The proposed technique aims to locate the significant areas in facial regions from which the significant features are extracted. experiments, the PCA projection method is applied to local facial regions instead of the whole face images to extract features. After all generations, the total number of times that each feature has been selected for the best classification rate is calculated. The best classification results for each facial region feature set When the size of the small rectangular area was 6 � 4, the hidden units 40 and 56 feature selections achieved the best recognition rate on Database 1. improving recognition rate compared to just one facial region feature set. Selection and fusion of facial features for face recognition Selection and fusion of facial features for face recognition Selection and fusion of facial features for face recognition	./cache/work_igv6r4iavzgn3mtitipb2naxam.pdf	./txt/work_igv6r4iavzgn3mtitipb2naxam.txt