A novel separability objective function in CNN for Feature Extraction of SAR Images

Fei Gao; Meng Wang; Jun Wang; Erfu Yang; Huiyu Zhou

doi:10.1049/cje.2018.12.001

A novel separability objective function in CNN for Feature Extraction of SAR Images

Fei Gao, Meng Wang, Jun Wang^*, Erfu Yang, Huiyu Zhou

^*Corresponding author for this work

Research output: Contribution to journal › Article › peer-review

2 Citations (Scopus)

Abstract

Convolutional neural network (CNN) has become a promising method for Synthetic aperture radar (SAR) target recognition. Existing CNN models aim at seeking the best separation between classes, but rarely care about the separability of them. We performs a separability measure by analyzing the property of linear separability, and proposes an objective function for CNN to extract linearly separable features. The experimental results indicate the output features are linearly separable, and the classification results are comparable with the other state of the art techniques.

Original language	English
Pages (from-to)	423-429
Number of pages	7
Journal	CHINESE JOURNAL OF ELECTRONICS
Volume	28
Issue number	2
DOIs	https://doi.org/10.1049/cje.2018.12.001
Publication status	Published - 01 Mar 2019
Externally published	Yes

Bibliographical note

Funding Information:
∗Manuscript Received Apr. 12, 2018; Accepted Aug. 11, 2018. This research was funded by the National Natural Science Foundation of China (No.61771027, No.61071139, No.61471019, No.61501011, No.61171122). E. Yang is supported in part under the RSE-NNSFC Joint Project (2017-2019) (No.6161101383) with China University of Petroleum (Huadong). H. Zhou is supported by Invest NI/Philips, UK EPSRC (No.EP/N011074/1) and Royal Society-Newton Advanced Fellowship (No.NA160342). © 2019 Chinese Institute of Electronics. DOI:10.1049/cje.2018.12.001

Funding Information:
This research was funded by the National Natural Science Foundation of China (No.61771027, No.61071139, No.61471019, No.61501011, No.61171122). E. Yang is supported in part under the RSE-NNSFC Joint Project (2017-2019) (No.6161101383) with China University of Petroleum (Huadong). H. Zhou is supported by Invest NI/Philips, UK EPSRC (No.EP/N011074/1) and Royal Society-Newton Advanced Fellowship (No.NA160342).

Publisher Copyright:
© 2019 Chinese Institute of Electronics.

Copyright:
Copyright 2020 Elsevier B.V., All rights reserved.

Keywords

Classification
Convolution neural network (CNN)
Linear separability
Objective function
Synthetic aperture radar (SAR)

ASJC Scopus subject areas

Electrical and Electronic Engineering
Applied Mathematics

Access to Document

10.1049/cje.2018.12.001Licence: Unspecified

Fingerprint Dive into the research topics of 'A novel separability objective function in CNN for Feature Extraction of SAR Images'. Together they form a unique fingerprint.

Cite this

@article{29ddb3d50e924c22a3e77f908cff372c,

title = "A novel separability objective function in CNN for Feature Extraction of SAR Images",

abstract = "Convolutional neural network (CNN) has become a promising method for Synthetic aperture radar (SAR) target recognition. Existing CNN models aim at seeking the best separation between classes, but rarely care about the separability of them. We performs a separability measure by analyzing the property of linear separability, and proposes an objective function for CNN to extract linearly separable features. The experimental results indicate the output features are linearly separable, and the classification results are comparable with the other state of the art techniques.",

keywords = "Classification, Convolution neural network (CNN), Linear separability, Objective function, Synthetic aperture radar (SAR)",

author = "Fei Gao and Meng Wang and Jun Wang and Erfu Yang and Huiyu Zhou",

note = "Funding Information: ∗Manuscript Received Apr. 12, 2018; Accepted Aug. 11, 2018. This research was funded by the National Natural Science Foundation of China (No.61771027, No.61071139, No.61471019, No.61501011, No.61171122). E. Yang is supported in part under the RSE-NNSFC Joint Project (2017-2019) (No.6161101383) with China University of Petroleum (Huadong). H. Zhou is supported by Invest NI/Philips, UK EPSRC (No.EP/N011074/1) and Royal Society-Newton Advanced Fellowship (No.NA160342). {\textcopyright} 2019 Chinese Institute of Electronics. DOI:10.1049/cje.2018.12.001 Funding Information: This research was funded by the National Natural Science Foundation of China (No.61771027, No.61071139, No.61471019, No.61501011, No.61171122). E. Yang is supported in part under the RSE-NNSFC Joint Project (2017-2019) (No.6161101383) with China University of Petroleum (Huadong). H. Zhou is supported by Invest NI/Philips, UK EPSRC (No.EP/N011074/1) and Royal Society-Newton Advanced Fellowship (No.NA160342). Publisher Copyright: {\textcopyright} 2019 Chinese Institute of Electronics. Copyright: Copyright 2020 Elsevier B.V., All rights reserved.",

year = "2019",

month = mar,

day = "1",

doi = "10.1049/cje.2018.12.001",

language = "English",

volume = "28",

pages = "423--429",

journal = "CHINESE JOURNAL OF ELECTRONICS",

issn = "1022-4653",

publisher = "Chinese Institute of Electronics",

number = "2",

}

TY - JOUR

T1 - A novel separability objective function in CNN for Feature Extraction of SAR Images

AU - Gao, Fei

AU - Wang, Meng

AU - Wang, Jun

AU - Yang, Erfu

AU - Zhou, Huiyu

N1 - Funding Information: ∗Manuscript Received Apr. 12, 2018; Accepted Aug. 11, 2018. This research was funded by the National Natural Science Foundation of China (No.61771027, No.61071139, No.61471019, No.61501011, No.61171122). E. Yang is supported in part under the RSE-NNSFC Joint Project (2017-2019) (No.6161101383) with China University of Petroleum (Huadong). H. Zhou is supported by Invest NI/Philips, UK EPSRC (No.EP/N011074/1) and Royal Society-Newton Advanced Fellowship (No.NA160342). © 2019 Chinese Institute of Electronics. DOI:10.1049/cje.2018.12.001 Funding Information: This research was funded by the National Natural Science Foundation of China (No.61771027, No.61071139, No.61471019, No.61501011, No.61171122). E. Yang is supported in part under the RSE-NNSFC Joint Project (2017-2019) (No.6161101383) with China University of Petroleum (Huadong). H. Zhou is supported by Invest NI/Philips, UK EPSRC (No.EP/N011074/1) and Royal Society-Newton Advanced Fellowship (No.NA160342). Publisher Copyright: © 2019 Chinese Institute of Electronics. Copyright: Copyright 2020 Elsevier B.V., All rights reserved.

PY - 2019/3/1

Y1 - 2019/3/1

N2 - Convolutional neural network (CNN) has become a promising method for Synthetic aperture radar (SAR) target recognition. Existing CNN models aim at seeking the best separation between classes, but rarely care about the separability of them. We performs a separability measure by analyzing the property of linear separability, and proposes an objective function for CNN to extract linearly separable features. The experimental results indicate the output features are linearly separable, and the classification results are comparable with the other state of the art techniques.

AB - Convolutional neural network (CNN) has become a promising method for Synthetic aperture radar (SAR) target recognition. Existing CNN models aim at seeking the best separation between classes, but rarely care about the separability of them. We performs a separability measure by analyzing the property of linear separability, and proposes an objective function for CNN to extract linearly separable features. The experimental results indicate the output features are linearly separable, and the classification results are comparable with the other state of the art techniques.

KW - Classification

KW - Convolution neural network (CNN)

KW - Linear separability

KW - Objective function

KW - Synthetic aperture radar (SAR)

U2 - 10.1049/cje.2018.12.001

DO - 10.1049/cje.2018.12.001

M3 - Article

AN - SCOPUS:85064182054

VL - 28

SP - 423

EP - 429

JO - CHINESE JOURNAL OF ELECTRONICS

JF - CHINESE JOURNAL OF ELECTRONICS

SN - 1022-4653

IS - 2

ER -