Parallelization of AdaBoost algorithm on multi-core processors

This paper examines and extracts the parallelism in the AdaBoost person detection algorithm on multi-core processors. As multi-core processors become pervasive, effectively executing many threads simultaneously is crucial in harnessing the computation power. Although the application exposes many levels of parallelism, none of them delivers a satisfactory scaling performance on newest multi-core processors due to load imbalance and parallel overhead. This paper demonstrates how to analyze the thread-level parallelism, and how to choose appropriate one to utilize current 4-core and 8-core processors. With careful optimization and parallelization, the AdaBoost person detection algorithm can efficiently utilize the power of multi-core processors, and now it is 7 times faster than the serial version.

[1]  Paul A. Viola,et al.  Rapid object detection using a boosted cascade of simple features , 2001, Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. CVPR 2001.

[2]  Bill Triggs,et al.  Histograms of oriented gradients for human detection , 2005, 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05).

[3]  Ming Yang,et al.  Face detection for automatic exposure control in handheld camera , 2006, Fourth IEEE International Conference on Computer Vision Systems (ICVS'06).

[4]  Saurabh Dighe,et al.  An 80-Tile 1.28TFLOPS Network-on-Chip in 65nm CMOS , 2007, 2007 IEEE International Solid-State Circuits Conference. Digest of Technical Papers.

[5]  Ivan Laptev,et al.  Improvements of Object Detection Using Boosted Histograms , 2006, BMVC.

[6]  Christopher J. Hughes,et al.  Carbon: architectural support for fine-grained parallelism on chip multiprocessors , 2007, ISCA '07.

[7]  Pradeep Dubey,et al.  Larrabee: A Many-Core x86 Architecture for Visual Computing , 2009, IEEE Micro.

[8]  Vinod Nair,et al.  An FPGA-Based People Detection System , 2005, EURASIP J. Adv. Signal Process..

[9]  Yu Wei,et al.  FPGA implementation of AdaBoost algorithm for detection of face biometrics , 2004, IEEE International Workshop on Biomedical Circuits and Systems, 2004..

[10]  Narayanan Vijaykrishnan,et al.  A parallel architecture for hardware face detection , 2006, IEEE Computer Society Annual Symposium on Emerging VLSI Technologies and Architectures (ISVLSI'06).