Real-time pedestrian detection based on A hierarchical two-stage Support Vector Machine

Kyoungwon Min, Haengseon Son, Yoonsik Choe, Yong Goo Kim

Research output: Chapter in Book/Report/Conference proceedingConference contribution

3 Citations (Scopus)

Abstract

This Paper presents an SVM (Support Vector Machine) based real-time pedestrian detection scheme for next-generation automotive vision applications. To meet the requirement of real-time detection with high accuracy, we designed the proposed system consisting of 2-stage hierarchical SVMs. In the proposed system, most of the input data are classified by the 1st stage linear SVM and only the inputs between positive and negative hyper-plane of the linear SVM are transferred to the 2nd stage non-linear SVM. This hierarchical 2-stage classifier can be suited for various systems via controlling the amount of data processed by the 2nd stage classifier, which trades off the detection accuracy and the required system resources. To make the proposed 2nd stage non-linear SVM further appropriate for various systems, a hyper-plane approximation technique by sample pruning has been adopted. By reducing the number of required SVs (Support Vectors) using this technique and controlling the amount of data processed via the 2 nd stage classifier, high precision non-linear SVM can be employed in the proposed real-time pedestrian detection system. Simulations using HOG (Histogram of Oriented Gradient) features and Daimler pedestrian dataset show the proposed system provides highly accurate classification results under the real-time constraint of application.

Original languageEnglish
Title of host publicationProceedings of the 2013 IEEE 8th Conference on Industrial Electronics and Applications, ICIEA 2013
Pages114-119
Number of pages6
DOIs
Publication statusPublished - 2013 Aug 19
Event2013 IEEE 8th Conference on Industrial Electronics and Applications, ICIEA 2013 - Melbourne, VIC, Australia
Duration: 2013 Jun 192013 Jun 21

Publication series

NameProceedings of the 2013 IEEE 8th Conference on Industrial Electronics and Applications, ICIEA 2013

Other

Other2013 IEEE 8th Conference on Industrial Electronics and Applications, ICIEA 2013
CountryAustralia
CityMelbourne, VIC
Period13/6/1913/6/21

Fingerprint

Support vector machines
Classifiers

All Science Journal Classification (ASJC) codes

  • Electrical and Electronic Engineering

Cite this

Min, K., Son, H., Choe, Y., & Kim, Y. G. (2013). Real-time pedestrian detection based on A hierarchical two-stage Support Vector Machine. In Proceedings of the 2013 IEEE 8th Conference on Industrial Electronics and Applications, ICIEA 2013 (pp. 114-119). [6566350] (Proceedings of the 2013 IEEE 8th Conference on Industrial Electronics and Applications, ICIEA 2013). https://doi.org/10.1109/ICIEA.2013.6566350
Min, Kyoungwon ; Son, Haengseon ; Choe, Yoonsik ; Kim, Yong Goo. / Real-time pedestrian detection based on A hierarchical two-stage Support Vector Machine. Proceedings of the 2013 IEEE 8th Conference on Industrial Electronics and Applications, ICIEA 2013. 2013. pp. 114-119 (Proceedings of the 2013 IEEE 8th Conference on Industrial Electronics and Applications, ICIEA 2013).
@inproceedings{def1b28cd69749c49c64a63ccd5646e4,
title = "Real-time pedestrian detection based on A hierarchical two-stage Support Vector Machine",
abstract = "This Paper presents an SVM (Support Vector Machine) based real-time pedestrian detection scheme for next-generation automotive vision applications. To meet the requirement of real-time detection with high accuracy, we designed the proposed system consisting of 2-stage hierarchical SVMs. In the proposed system, most of the input data are classified by the 1st stage linear SVM and only the inputs between positive and negative hyper-plane of the linear SVM are transferred to the 2nd stage non-linear SVM. This hierarchical 2-stage classifier can be suited for various systems via controlling the amount of data processed by the 2nd stage classifier, which trades off the detection accuracy and the required system resources. To make the proposed 2nd stage non-linear SVM further appropriate for various systems, a hyper-plane approximation technique by sample pruning has been adopted. By reducing the number of required SVs (Support Vectors) using this technique and controlling the amount of data processed via the 2 nd stage classifier, high precision non-linear SVM can be employed in the proposed real-time pedestrian detection system. Simulations using HOG (Histogram of Oriented Gradient) features and Daimler pedestrian dataset show the proposed system provides highly accurate classification results under the real-time constraint of application.",
author = "Kyoungwon Min and Haengseon Son and Yoonsik Choe and Kim, {Yong Goo}",
year = "2013",
month = "8",
day = "19",
doi = "10.1109/ICIEA.2013.6566350",
language = "English",
isbn = "9781467363211",
series = "Proceedings of the 2013 IEEE 8th Conference on Industrial Electronics and Applications, ICIEA 2013",
pages = "114--119",
booktitle = "Proceedings of the 2013 IEEE 8th Conference on Industrial Electronics and Applications, ICIEA 2013",

}

Min, K, Son, H, Choe, Y & Kim, YG 2013, Real-time pedestrian detection based on A hierarchical two-stage Support Vector Machine. in Proceedings of the 2013 IEEE 8th Conference on Industrial Electronics and Applications, ICIEA 2013., 6566350, Proceedings of the 2013 IEEE 8th Conference on Industrial Electronics and Applications, ICIEA 2013, pp. 114-119, 2013 IEEE 8th Conference on Industrial Electronics and Applications, ICIEA 2013, Melbourne, VIC, Australia, 13/6/19. https://doi.org/10.1109/ICIEA.2013.6566350

Real-time pedestrian detection based on A hierarchical two-stage Support Vector Machine. / Min, Kyoungwon; Son, Haengseon; Choe, Yoonsik; Kim, Yong Goo.

Proceedings of the 2013 IEEE 8th Conference on Industrial Electronics and Applications, ICIEA 2013. 2013. p. 114-119 6566350 (Proceedings of the 2013 IEEE 8th Conference on Industrial Electronics and Applications, ICIEA 2013).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Real-time pedestrian detection based on A hierarchical two-stage Support Vector Machine

AU - Min, Kyoungwon

AU - Son, Haengseon

AU - Choe, Yoonsik

AU - Kim, Yong Goo

PY - 2013/8/19

Y1 - 2013/8/19

N2 - This Paper presents an SVM (Support Vector Machine) based real-time pedestrian detection scheme for next-generation automotive vision applications. To meet the requirement of real-time detection with high accuracy, we designed the proposed system consisting of 2-stage hierarchical SVMs. In the proposed system, most of the input data are classified by the 1st stage linear SVM and only the inputs between positive and negative hyper-plane of the linear SVM are transferred to the 2nd stage non-linear SVM. This hierarchical 2-stage classifier can be suited for various systems via controlling the amount of data processed by the 2nd stage classifier, which trades off the detection accuracy and the required system resources. To make the proposed 2nd stage non-linear SVM further appropriate for various systems, a hyper-plane approximation technique by sample pruning has been adopted. By reducing the number of required SVs (Support Vectors) using this technique and controlling the amount of data processed via the 2 nd stage classifier, high precision non-linear SVM can be employed in the proposed real-time pedestrian detection system. Simulations using HOG (Histogram of Oriented Gradient) features and Daimler pedestrian dataset show the proposed system provides highly accurate classification results under the real-time constraint of application.

AB - This Paper presents an SVM (Support Vector Machine) based real-time pedestrian detection scheme for next-generation automotive vision applications. To meet the requirement of real-time detection with high accuracy, we designed the proposed system consisting of 2-stage hierarchical SVMs. In the proposed system, most of the input data are classified by the 1st stage linear SVM and only the inputs between positive and negative hyper-plane of the linear SVM are transferred to the 2nd stage non-linear SVM. This hierarchical 2-stage classifier can be suited for various systems via controlling the amount of data processed by the 2nd stage classifier, which trades off the detection accuracy and the required system resources. To make the proposed 2nd stage non-linear SVM further appropriate for various systems, a hyper-plane approximation technique by sample pruning has been adopted. By reducing the number of required SVs (Support Vectors) using this technique and controlling the amount of data processed via the 2 nd stage classifier, high precision non-linear SVM can be employed in the proposed real-time pedestrian detection system. Simulations using HOG (Histogram of Oriented Gradient) features and Daimler pedestrian dataset show the proposed system provides highly accurate classification results under the real-time constraint of application.

UR - http://www.scopus.com/inward/record.url?scp=84881407509&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84881407509&partnerID=8YFLogxK

U2 - 10.1109/ICIEA.2013.6566350

DO - 10.1109/ICIEA.2013.6566350

M3 - Conference contribution

AN - SCOPUS:84881407509

SN - 9781467363211

T3 - Proceedings of the 2013 IEEE 8th Conference on Industrial Electronics and Applications, ICIEA 2013

SP - 114

EP - 119

BT - Proceedings of the 2013 IEEE 8th Conference on Industrial Electronics and Applications, ICIEA 2013

ER -

Min K, Son H, Choe Y, Kim YG. Real-time pedestrian detection based on A hierarchical two-stage Support Vector Machine. In Proceedings of the 2013 IEEE 8th Conference on Industrial Electronics and Applications, ICIEA 2013. 2013. p. 114-119. 6566350. (Proceedings of the 2013 IEEE 8th Conference on Industrial Electronics and Applications, ICIEA 2013). https://doi.org/10.1109/ICIEA.2013.6566350