Fast LBP Face Detection on Low-Power SIMD Architectures

Author

Bilaniuk, Olexa ; Fazl-Ersi, Ehsan ; Laganiere, Robert ; Xu, Changsheng ; Laroche, Daniel ; Moulder, Craig

Author_Institution

Univ. of Ottawa, Ottawa, ON, Canada

fYear

2014

fDate

23-28 June 2014

Firstpage

630

Lastpage

636

Abstract

This paper presents an embedded implementation of a face detection method based on boosted LBP features for Single Instruction Multiple Data (SIMD) architectures. The implementation exploits parallelism and data reuse in the detection algorithm and is integrated into CogniVue´s Gen-1 APEX platform, which uses a SIMD design and is extremely energy efficient. The proposed embedded face detection system runs at 5 VGA frames per second, while providing similar accuracy to the PC version of the LBP face detection algorithm included in the OpenCV library.

Keywords

embedded systems; object detection; parallel processing; power aware computing; CogniVue Gen-1 APEX platform; LBP face detection; OpenCV library; data reuse; embedded face detection system; energy efficiency; local binary pattern; low-power SIMD architectures; single instruction multiple data; Face; Face detection; Feature extraction; Field programmable gate arrays; Table lookup; Vectors; LBP features; SIMD architecture; face detection;

fLanguage

English

Publisher

ieee

Conference_Titel

Computer Vision and Pattern Recognition Workshops (CVPRW), 2014 IEEE Conference on

Conference_Location

Columbus, OH

Type

conf

DOI

10.1109/CVPRW.2014.96

Filename

6910046