Pyramidal channel features for pedestrian detector

Young Chul Lim; Minsung Kang

doi:10.1145/2811411.2811511

Pyramidal channel features for pedestrian detector

Young Chul Lim
, Minsung Kang

Division of Automotive Technology

Daegu Gyeongbuk Institute of Science and Technology

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

1 Scopus citations

Abstract

Most state-of-the-art pedestrian detectors diversify features to improve the detection performance. Among such features, the gradient histogram is the most informative. This work investigates how diverse orientation information influences the pedestrian detection performance. In this paper, we propose a pedestrian detection method which uses pyramidal aggregated channel features which are generated with a layer-by-layer assembly scheme. Experimental results on the INRIA dataset show that our method can reach state-of-the-art performance despite using only gradient features.

Original language	English
Title of host publication	Proceeding of the 2015 Research in Adaptive and Convergent Systems, RACS 2015
Publisher	Association for Computing Machinery, Inc
Pages	195-199
Number of pages	5
ISBN (Electronic)	9781450337380
DOIs	https://doi.org/10.1145/2811411.2811511
State	Published - 9 Oct 2015
Event	Research in Adaptive and Convergent Systems, RACS 2015 - Prague, Czech Republic Duration: 9 Oct 2015 → 12 Oct 2015

Publication series

Name	Proceeding of the 2015 Research in Adaptive and Convergent Systems, RACS 2015

Conference

Conference	Research in Adaptive and Convergent Systems, RACS 2015
Country/Territory	Czech Republic
City	Prague
Period	9/10/15 → 12/10/15

Bibliographical note

Publisher Copyright:
© 2015 ACM.

Keywords

Feature extraction
Gradient histogram
Pedestrian detection

Access to Document

10.1145/2811411.2811511

Cite this

@inproceedings{481ba8d7cdfb47dca21d6039f0b7591f,

title = "Pyramidal channel features for pedestrian detector",

abstract = "Most state-of-the-art pedestrian detectors diversify features to improve the detection performance. Among such features, the gradient histogram is the most informative. This work investigates how diverse orientation information influences the pedestrian detection performance. In this paper, we propose a pedestrian detection method which uses pyramidal aggregated channel features which are generated with a layer-by-layer assembly scheme. Experimental results on the INRIA dataset show that our method can reach state-of-the-art performance despite using only gradient features.",

keywords = "Feature extraction, Gradient histogram, Pedestrian detection",

author = "Lim, \{Young Chul\} and Minsung Kang",

note = "Publisher Copyright: {\textcopyright} 2015 ACM.; Research in Adaptive and Convergent Systems, RACS 2015 ; Conference date: 09-10-2015 Through 12-10-2015",

year = "2015",

month = oct,

day = "9",

doi = "10.1145/2811411.2811511",

language = "English",

series = "Proceeding of the 2015 Research in Adaptive and Convergent Systems, RACS 2015",

publisher = "Association for Computing Machinery, Inc",

pages = "195--199",

booktitle = "Proceeding of the 2015 Research in Adaptive and Convergent Systems, RACS 2015",

}

Lim, YC & Kang, M 2015, Pyramidal channel features for pedestrian detector. in Proceeding of the 2015 Research in Adaptive and Convergent Systems, RACS 2015. Proceeding of the 2015 Research in Adaptive and Convergent Systems, RACS 2015, Association for Computing Machinery, Inc, pp. 195-199, Research in Adaptive and Convergent Systems, RACS 2015, Prague, Czech Republic, 9/10/15. https://doi.org/10.1145/2811411.2811511

Pyramidal channel features for pedestrian detector. / Lim, Young Chul; Kang, Minsung.
Proceeding of the 2015 Research in Adaptive and Convergent Systems, RACS 2015. Association for Computing Machinery, Inc, 2015. p. 195-199 (Proceeding of the 2015 Research in Adaptive and Convergent Systems, RACS 2015).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Pyramidal channel features for pedestrian detector

AU - Lim, Young Chul

AU - Kang, Minsung

PY - 2015/10/9

Y1 - 2015/10/9

N2 - Most state-of-the-art pedestrian detectors diversify features to improve the detection performance. Among such features, the gradient histogram is the most informative. This work investigates how diverse orientation information influences the pedestrian detection performance. In this paper, we propose a pedestrian detection method which uses pyramidal aggregated channel features which are generated with a layer-by-layer assembly scheme. Experimental results on the INRIA dataset show that our method can reach state-of-the-art performance despite using only gradient features.

AB - Most state-of-the-art pedestrian detectors diversify features to improve the detection performance. Among such features, the gradient histogram is the most informative. This work investigates how diverse orientation information influences the pedestrian detection performance. In this paper, we propose a pedestrian detection method which uses pyramidal aggregated channel features which are generated with a layer-by-layer assembly scheme. Experimental results on the INRIA dataset show that our method can reach state-of-the-art performance despite using only gradient features.

KW - Feature extraction

KW - Gradient histogram

KW - Pedestrian detection

UR - https://www.scopus.com/pages/publications/84960899990

U2 - 10.1145/2811411.2811511

DO - 10.1145/2811411.2811511

M3 - Conference contribution

AN - SCOPUS:84960899990

T3 - Proceeding of the 2015 Research in Adaptive and Convergent Systems, RACS 2015

SP - 195

EP - 199

BT - Proceeding of the 2015 Research in Adaptive and Convergent Systems, RACS 2015

PB - Association for Computing Machinery, Inc

T2 - Research in Adaptive and Convergent Systems, RACS 2015

Y2 - 9 October 2015 through 12 October 2015

ER -

Pyramidal channel features for pedestrian detector

Abstract

Publication series

Conference

Bibliographical note

Keywords

Access to Document

Other files and links

Fingerprint

Cite this