Applied aspects in informatics
N.A. Bocharov, E.E. Limonova, B.N. Paramonov, S.A. Usilin Optimization for the Elbrus computing architecture of the modified method of Viola and Jones
Mathematical models of socio-economic processes
Dynamic systems
Scientometrics and management science
Recognition of images
N.A. Bocharov, E.E. Limonova, B.N. Paramonov, S.A. Usilin Optimization for the Elbrus computing architecture of the modified method of Viola and Jones

Abstract.

In the work, modern modifications of the original training algorithm by Viola and Jones are considered. A modified cascade learning algorithm for the Viola and Jones algorithm is proposed, based on the new structure of a high-level classifier. The proposed algorithm allows to increase the accuracy of recognition, and also provides the possibility of additional training. The optimization of the implemented improved algorithm of Viola and Jones for the Elbrus architecture is described, which allows to significantly increase the speed of the implemented algorithm for the Elbrus architecture.

Keywords:

object detection, Viola and Jones method, Elbrus architecture, classification algorithm.

PP. 12-23.

References

1. Huang C. et al. Vector boosting for rotation invariant multi-view face detection // Computer Vision, 2005. ICCV 2005. Tenth  IEEE International Conference on. 2005. Vol. 1. 446–453.
2. Li S.Z., Zhang Z.Q. FloatBoost learning and statistical face detection // IEEE Trans. Pattern Anal. Mach. Intell. 2004. Vol. 26, № 9. 1112–1123.
3. Domingo C., Watanabe O. MadaBoost: A Modification of AdaBoost // Conference on Computational Learning Theory (COLT). 2000. 180–189.
4. Xiao R., Zhu L., Zhang H.-J. Boosting Chain Learning for Object Detection // ICCV ’03: Proceedings of the Ninth IEEE International Conference on Computer Vision. 2003. 709. P.
5. Wu B. et al. Fast rotation invariant multi-view face detection based on real AdaBoost // In Sixth IEEE International Conference on Automatic Face and Gesture Recognition. 2004. P. 79–84.
6. Dollár P. et al. Integral Channel Features // BMVC 2009 London Engl. 2009. 1–11.
7. Bourdev L., Brandt J. Robust Object Detection via Soft Cascade // Proceedings of the 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR’05) - Volume 2 - Volume 02. Washington, DC, USA: IEEE  Computer Society, 2005. 236–243.
8. Wu B. et al. Fast rotation invariant multi-view face detection based on real AdaBoost // In Sixth IEEE International Conference on Automatic Face and Gesture Recognition. 2004. 79–84.
9. Huang C.H.C. et al. Vector boosting for rotation invariant multi-view face detection // Tenth IEEE Int. Conf. Comput. Vis. Vol. 1. 2005. Vol. 1.
10. Norushis A. Construction of logical (tree-like) classifiers of methods of top-down search (review) // Statistical problems of management. Vilnius, 1990. Vol. 93. 131-158.
11. Powers D.M.W. Evaluation: From Precision, Recall and F-Measure To Roc, Informedness, Markedness & Correlation // J.  Mach. Learn. Technol. 2011. Vol. 2, № 1. 37–63.
12. Everingham M. et al. The pascal visual object classes (VOC) challenge // Int. J. Comput. Vis. 2010. Vol. 88, № 2. 303–338.
13. Kim A.K., Bychkov I.N. Russian Technologies “Elbrus” for personal computers, servers and supercomputers // Modern  Information Technologies and IT Education, Moscow: Foundation for the Promotion of Internet Media, IT Education, Human  Capacity “League of Internet Media”, 2014, No. 10. 39-50.
14. Kim A.K., Bychkov I.N., Ermakov S.G. Microprocessors and computing systems of the Elbrus family. - SPb: Peter, 2013. -  272 С.
15. Ishin P.A., Loginov V.E., Vasilyev P.P. Acceleration of computations using highperformance mathematical and multimedia libraries for the architecture of Elbrus // Bulletin of Aerospace Defense, Moscow: Almaz Scientific and Production Association.  acad. A.A. Raspletin, 2015, No. 4 (8). 64-68.
 

2019-69-2
2019-69-1
2018-68-4
2018-S1

© ФИЦ ИУ РАН 2008-2018. Создание сайта "РосИнтернет технологии".