End-to-end pedestrian collision warning system based on a convolutional neural network with semantic segmentation

Heechul Jung; Min Kook Choi; Kwon Soon; Woo Young Jung

doi:10.1109/ICCE.2018.8326129

End-to-end pedestrian collision warning system based on a convolutional neural network with semantic segmentation

Heechul Jung
, Min Kook Choi
, Kwon Soon
, Woo Young Jung

Division of Automotive Technology

Daegu Gyeongbuk Institute of Science and Technology

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

11 Scopus citations

Abstract

Traditional pedestrian collision warning systems sometimes raise alarms even when there is no danger (e.g., when all pedestrians are walking on the sidewalk). These false alarms can make it difficult for drivers to concentrate on their driving. In this paper, we propose a novel framework for an end-to-end pedestrian collision warning system based on a convolutional neural network. Semantic segmentation information is used to train the convolutional neural network and two loss functions, such as cross entropy and Euclidean losses, are minimized. Finally, we demonstrate the effectiveness of our method in reducing false alarms and increasing warning accuracy compared to a traditional histogram of oriented gradients (HOG)-based system.

Original language	English
Title of host publication	2018 IEEE International Conference on Consumer Electronics, ICCE 2018
Editors	Saraju P. Mohanty, Peter Corcoran, Hai Li, Anirban Sengupta, Jong-Hyouk Lee
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	1-3
Number of pages	3
ISBN (Electronic)	9781538630259
DOIs	https://doi.org/10.1109/ICCE.2018.8326129
State	Published - 26 Mar 2018
Event	2018 IEEE International Conference on Consumer Electronics, ICCE 2018 - Las Vegas, United States Duration: 12 Jan 2018 → 14 Jan 2018

Publication series

Name	2018 IEEE International Conference on Consumer Electronics, ICCE 2018
Volume	2018-January

Conference

Conference	2018 IEEE International Conference on Consumer Electronics, ICCE 2018
Country/Territory	United States
City	Las Vegas
Period	12/01/18 → 14/01/18

Bibliographical note

Publisher Copyright:
© 2018 IEEE.

Access to Document

10.1109/ICCE.2018.8326129

Cite this

Jung, H., Choi, M. K., Soon, K., & Jung, W. Y. (2018). End-to-end pedestrian collision warning system based on a convolutional neural network with semantic segmentation. In S. P. Mohanty, P. Corcoran, H. Li, A. Sengupta, & J.-H. Lee (Eds.), 2018 IEEE International Conference on Consumer Electronics, ICCE 2018 (pp. 1-3). (2018 IEEE International Conference on Consumer Electronics, ICCE 2018; Vol. 2018-January). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ICCE.2018.8326129

Jung, Heechul ; Choi, Min Kook ; Soon, Kwon et al. / End-to-end pedestrian collision warning system based on a convolutional neural network with semantic segmentation. 2018 IEEE International Conference on Consumer Electronics, ICCE 2018. editor / Saraju P. Mohanty ; Peter Corcoran ; Hai Li ; Anirban Sengupta ; Jong-Hyouk Lee. Institute of Electrical and Electronics Engineers Inc., 2018. pp. 1-3 (2018 IEEE International Conference on Consumer Electronics, ICCE 2018).

@inproceedings{265156391fc941ecb47b07eee01d1d85,

title = "End-to-end pedestrian collision warning system based on a convolutional neural network with semantic segmentation",

abstract = "Traditional pedestrian collision warning systems sometimes raise alarms even when there is no danger (e.g., when all pedestrians are walking on the sidewalk). These false alarms can make it difficult for drivers to concentrate on their driving. In this paper, we propose a novel framework for an end-to-end pedestrian collision warning system based on a convolutional neural network. Semantic segmentation information is used to train the convolutional neural network and two loss functions, such as cross entropy and Euclidean losses, are minimized. Finally, we demonstrate the effectiveness of our method in reducing false alarms and increasing warning accuracy compared to a traditional histogram of oriented gradients (HOG)-based system.",

author = "Heechul Jung and Choi, \{Min Kook\} and Kwon Soon and Jung, \{Woo Young\}",

note = "Publisher Copyright: {\textcopyright} 2018 IEEE.; 2018 IEEE International Conference on Consumer Electronics, ICCE 2018 ; Conference date: 12-01-2018 Through 14-01-2018",

year = "2018",

month = mar,

day = "26",

doi = "10.1109/ICCE.2018.8326129",

language = "English",

series = "2018 IEEE International Conference on Consumer Electronics, ICCE 2018",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "1--3",

editor = "Mohanty, \{Saraju P.\} and Peter Corcoran and Hai Li and Anirban Sengupta and Jong-Hyouk Lee",

booktitle = "2018 IEEE International Conference on Consumer Electronics, ICCE 2018",

}

Jung, H, Choi, MK, Soon, K & Jung, WY 2018, End-to-end pedestrian collision warning system based on a convolutional neural network with semantic segmentation. in SP Mohanty, P Corcoran, H Li, A Sengupta & J-H Lee (eds), 2018 IEEE International Conference on Consumer Electronics, ICCE 2018. 2018 IEEE International Conference on Consumer Electronics, ICCE 2018, vol. 2018-January, Institute of Electrical and Electronics Engineers Inc., pp. 1-3, 2018 IEEE International Conference on Consumer Electronics, ICCE 2018, Las Vegas, United States, 12/01/18. https://doi.org/10.1109/ICCE.2018.8326129

End-to-end pedestrian collision warning system based on a convolutional neural network with semantic segmentation. / Jung, Heechul; Choi, Min Kook; Soon, Kwon et al.
2018 IEEE International Conference on Consumer Electronics, ICCE 2018. ed. / Saraju P. Mohanty; Peter Corcoran; Hai Li; Anirban Sengupta; Jong-Hyouk Lee. Institute of Electrical and Electronics Engineers Inc., 2018. p. 1-3 (2018 IEEE International Conference on Consumer Electronics, ICCE 2018; Vol. 2018-January).

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

TY - GEN

T1 - End-to-end pedestrian collision warning system based on a convolutional neural network with semantic segmentation

AU - Jung, Heechul

AU - Choi, Min Kook

AU - Soon, Kwon

AU - Jung, Woo Young

PY - 2018/3/26

Y1 - 2018/3/26

N2 - Traditional pedestrian collision warning systems sometimes raise alarms even when there is no danger (e.g., when all pedestrians are walking on the sidewalk). These false alarms can make it difficult for drivers to concentrate on their driving. In this paper, we propose a novel framework for an end-to-end pedestrian collision warning system based on a convolutional neural network. Semantic segmentation information is used to train the convolutional neural network and two loss functions, such as cross entropy and Euclidean losses, are minimized. Finally, we demonstrate the effectiveness of our method in reducing false alarms and increasing warning accuracy compared to a traditional histogram of oriented gradients (HOG)-based system.

AB - Traditional pedestrian collision warning systems sometimes raise alarms even when there is no danger (e.g., when all pedestrians are walking on the sidewalk). These false alarms can make it difficult for drivers to concentrate on their driving. In this paper, we propose a novel framework for an end-to-end pedestrian collision warning system based on a convolutional neural network. Semantic segmentation information is used to train the convolutional neural network and two loss functions, such as cross entropy and Euclidean losses, are minimized. Finally, we demonstrate the effectiveness of our method in reducing false alarms and increasing warning accuracy compared to a traditional histogram of oriented gradients (HOG)-based system.

UR - https://www.scopus.com/pages/publications/85048750464

U2 - 10.1109/ICCE.2018.8326129

DO - 10.1109/ICCE.2018.8326129

M3 - Conference contribution

AN - SCOPUS:85048750464

T3 - 2018 IEEE International Conference on Consumer Electronics, ICCE 2018

SP - 1

EP - 3

BT - 2018 IEEE International Conference on Consumer Electronics, ICCE 2018

A2 - Mohanty, Saraju P.

A2 - Corcoran, Peter

A2 - Li, Hai

A2 - Sengupta, Anirban

A2 - Lee, Jong-Hyouk

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 2018 IEEE International Conference on Consumer Electronics, ICCE 2018

Y2 - 12 January 2018 through 14 January 2018

ER -

Jung H, Choi MK, Soon K, Jung WY. End-to-end pedestrian collision warning system based on a convolutional neural network with semantic segmentation. In Mohanty SP, Corcoran P, Li H, Sengupta A, Lee JH, editors, 2018 IEEE International Conference on Consumer Electronics, ICCE 2018. Institute of Electrical and Electronics Engineers Inc. 2018. p. 1-3. (2018 IEEE International Conference on Consumer Electronics, ICCE 2018). doi: 10.1109/ICCE.2018.8326129

End-to-end pedestrian collision warning system based on a convolutional neural network with semantic segmentation

Abstract

Publication series

Conference

Bibliographical note

Access to Document

Other files and links

Fingerprint

Cite this