A hierarchical Float-Boost and MLP classifier for mobile phone embedded eye location system

CORC > 沈阳自动化研究所 > 中国科学院沈阳自动化研究所 > 机器人学研究室

	A hierarchical Float-Boost and MLP classifier for mobile phone embedded eye location system
	Chen D(陈丹); Tang XS(唐旭晟); Ou ZY(欧宗瑛); Xi N(席宁)
	2006
会议名称	3rd International Symposium on Neural Networks (ISNN 2006)
会议日期	May 28-31, 2006
会议地点	Chengdu, China
页码	20-25
中文摘要	This paper is focused on cellular phone embedded eye location system. The proposed eye detection system is based on a hierarchy cascade Float-Boost classifier combined with an MLP neural net post classifier. The system firstly locates the face and eye candidates' areas in the whole image by a hierarchical Float-Boost classifier. Then geometrical and relative position information of eye-pair and the face are extracted. These features are input to a MLP neural net post classier to arrive at an eye/non-eye decision. Experimental results show that our cellular phone embedded eye detection system can accurately locate double eyes with less computational and memory cost. It runs at 400ms per image of size 256x256 pixels with high detection rates on a SANYO cellular phone with ARM926EJ-S processor that lacks floating-point hardware.
收录类别	SCI ; EI ; CPCI（ISTP）
产权排序	1
会议主办者	Univ Electr Sci & Technol China, Chinese Univ Hong Kong, Asia Pacific Neural Network Assembly, European Neural Network Soc, IEEE Circuits & Syst Soc, IEEE Computat Intelligence Soc, Int Neural Network Soc, Natl Nat Sci Fdn China, KC Wong Educ Fdn Hong Kong
会议录	ADVANCES IN NEURAL NETWORKS - ISNN 2006, PT 2, PROCEEDINGS
会议录出版者	SPRINGER-VERLAG
会议录出版地	BERLIN
语种	英语
ISSN号	0302-9743
ISBN号	3-540-34437-3
WOS记录号	WOS:000239483000004
研究领域[WOS]	Computer Science
WOS标题词	Science & Technology ; Technology
内容类型	会议论文
源URL	[http://ir.sia.cn/handle/173321/8477]
专题	沈阳自动化研究所_机器人学研究室
推荐引用方式 GB/T 7714	Chen D,Tang XS,Ou ZY,et al. A hierarchical Float-Boost and MLP classifier for mobile phone embedded eye location system[C]. 见:3rd International Symposium on Neural Networks (ISNN 2006). Chengdu, China. May 28-31, 2006.