Background subtraction using encoder-decoder structured convolutional neural network

Kyungsun Lim, Won Dong Jang, Chang-Su Kim

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

11 Citations (Scopus)

Abstract

A background subtraction algorithm using an encoderdecoder structured convolutional neural network is proposed in this work, in order to segment out moving objects from the background. A target frame, its previous frame, and a background model are concatenated and fed into the network as the input. Then, the encoder generates a highlevel feature vector, and the decoder converts the feature vector into a segmentation map, which roughly identifies moving object regions. Moreover, we develop background modeling and foreground extraction techniques, which exploit contour information. Experimental results on the CD-net2014 dataset demonstrate that the proposed algorithm outperforms state-of-the-art techniques significantly.

Original languageEnglish
Title of host publication2017 14th IEEE International Conference on Advanced Video and Signal Based Surveillance, AVSS 2017
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9781538629390
DOIs
Publication statusPublished - 2017 Oct 20
Event14th IEEE International Conference on Advanced Video and Signal Based Surveillance, AVSS 2017 - Lecce, Italy
Duration: 2017 Aug 292017 Sep 1

Other

Other14th IEEE International Conference on Advanced Video and Signal Based Surveillance, AVSS 2017
CountryItaly
CityLecce
Period17/8/2917/9/1

Fingerprint

Neural networks

ASJC Scopus subject areas

  • Computer Vision and Pattern Recognition
  • Signal Processing

Cite this

Lim, K., Jang, W. D., & Kim, C-S. (2017). Background subtraction using encoder-decoder structured convolutional neural network. In 2017 14th IEEE International Conference on Advanced Video and Signal Based Surveillance, AVSS 2017 [8078547] Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/AVSS.2017.8078547

Background subtraction using encoder-decoder structured convolutional neural network. / Lim, Kyungsun; Jang, Won Dong; Kim, Chang-Su.

2017 14th IEEE International Conference on Advanced Video and Signal Based Surveillance, AVSS 2017. Institute of Electrical and Electronics Engineers Inc., 2017. 8078547.

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

Lim, K, Jang, WD & Kim, C-S 2017, Background subtraction using encoder-decoder structured convolutional neural network. in 2017 14th IEEE International Conference on Advanced Video and Signal Based Surveillance, AVSS 2017., 8078547, Institute of Electrical and Electronics Engineers Inc., 14th IEEE International Conference on Advanced Video and Signal Based Surveillance, AVSS 2017, Lecce, Italy, 17/8/29. https://doi.org/10.1109/AVSS.2017.8078547
Lim K, Jang WD, Kim C-S. Background subtraction using encoder-decoder structured convolutional neural network. In 2017 14th IEEE International Conference on Advanced Video and Signal Based Surveillance, AVSS 2017. Institute of Electrical and Electronics Engineers Inc. 2017. 8078547 https://doi.org/10.1109/AVSS.2017.8078547
Lim, Kyungsun ; Jang, Won Dong ; Kim, Chang-Su. / Background subtraction using encoder-decoder structured convolutional neural network. 2017 14th IEEE International Conference on Advanced Video and Signal Based Surveillance, AVSS 2017. Institute of Electrical and Electronics Engineers Inc., 2017.
@inproceedings{ae72fc5489c74f8a9041b3d360fce5fe,
title = "Background subtraction using encoder-decoder structured convolutional neural network",
abstract = "A background subtraction algorithm using an encoderdecoder structured convolutional neural network is proposed in this work, in order to segment out moving objects from the background. A target frame, its previous frame, and a background model are concatenated and fed into the network as the input. Then, the encoder generates a highlevel feature vector, and the decoder converts the feature vector into a segmentation map, which roughly identifies moving object regions. Moreover, we develop background modeling and foreground extraction techniques, which exploit contour information. Experimental results on the CD-net2014 dataset demonstrate that the proposed algorithm outperforms state-of-the-art techniques significantly.",
author = "Kyungsun Lim and Jang, {Won Dong} and Chang-Su Kim",
year = "2017",
month = "10",
day = "20",
doi = "10.1109/AVSS.2017.8078547",
language = "English",
booktitle = "2017 14th IEEE International Conference on Advanced Video and Signal Based Surveillance, AVSS 2017",
publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - GEN

T1 - Background subtraction using encoder-decoder structured convolutional neural network

AU - Lim, Kyungsun

AU - Jang, Won Dong

AU - Kim, Chang-Su

PY - 2017/10/20

Y1 - 2017/10/20

N2 - A background subtraction algorithm using an encoderdecoder structured convolutional neural network is proposed in this work, in order to segment out moving objects from the background. A target frame, its previous frame, and a background model are concatenated and fed into the network as the input. Then, the encoder generates a highlevel feature vector, and the decoder converts the feature vector into a segmentation map, which roughly identifies moving object regions. Moreover, we develop background modeling and foreground extraction techniques, which exploit contour information. Experimental results on the CD-net2014 dataset demonstrate that the proposed algorithm outperforms state-of-the-art techniques significantly.

AB - A background subtraction algorithm using an encoderdecoder structured convolutional neural network is proposed in this work, in order to segment out moving objects from the background. A target frame, its previous frame, and a background model are concatenated and fed into the network as the input. Then, the encoder generates a highlevel feature vector, and the decoder converts the feature vector into a segmentation map, which roughly identifies moving object regions. Moreover, we develop background modeling and foreground extraction techniques, which exploit contour information. Experimental results on the CD-net2014 dataset demonstrate that the proposed algorithm outperforms state-of-the-art techniques significantly.

UR - http://www.scopus.com/inward/record.url?scp=85039918235&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85039918235&partnerID=8YFLogxK

U2 - 10.1109/AVSS.2017.8078547

DO - 10.1109/AVSS.2017.8078547

M3 - Conference contribution

BT - 2017 14th IEEE International Conference on Advanced Video and Signal Based Surveillance, AVSS 2017

PB - Institute of Electrical and Electronics Engineers Inc.

ER -