Low-Complexity Fixed-Point Convolutional Neural Networks for Automatic Target Recognition

Hassan Dbouk; Hanfei Geng; Craig M. Vineyard; Naresh R. Shanbhag

doi:10.1109/ICASSP40776.2020.9054094

Low-Complexity Fixed-Point Convolutional Neural Networks for Automatic Target Recognition

Hassan Dbouk, Hanfei Geng, Craig M. Vineyard, Naresh R. Shanbhag

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

Abstract

There has been growing interest in developing neural network based automatic target recognition systems for synthetic aperture radar applications. However, these networks are typically complex in terms of storage and computation which inhibits their deployment in the field, where such resources are heavily constrained. In order to bring the cost of implementing these networks down, we develop a set of compact network architectures and train them in fixed-point. Our proposed method achieves an overall 984 reduction in terms of storage requirements and 71 × reduction in terms of computational complexity compared to state-of-the-art con-volutional neural networks for automatic target recognition (ATR), while maintaining a classification accuracy of > 99% on the MSTAR dataset.

Original language	English (US)
Title of host publication	2020 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2020 - Proceedings
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	1598-1602
Number of pages	5
ISBN (Electronic)	9781509066315
DOIs	https://doi.org/10.1109/ICASSP40776.2020.9054094
State	Published - May 2020
Event	2020 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2020 - Barcelona, Spain Duration: May 4 2020 → May 8 2020

Publication series

Name	ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings
Volume	2020-May
ISSN (Print)	1520-6149

Conference

Conference	2020 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2020
Country/Territory	Spain
City	Barcelona
Period	5/4/20 → 5/8/20

Keywords

automatic target recognition
deep learning
neural networks
quantization
synthetic aperture radar

ASJC Scopus subject areas

Software
Signal Processing
Electrical and Electronic Engineering

Online availability

10.1109/ICASSP40776.2020.9054094

Library availability

Discover UIUC Full Text

Cite this

Dbouk, H., Geng, H., Vineyard, C. M., & Shanbhag, N. R. (2020). Low-Complexity Fixed-Point Convolutional Neural Networks for Automatic Target Recognition. In 2020 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2020 - Proceedings (pp. 1598-1602). Article 9054094 (ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings; Vol. 2020-May). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ICASSP40776.2020.9054094

Low-Complexity Fixed-Point Convolutional Neural Networks for Automatic Target Recognition. / Dbouk, Hassan; Geng, Hanfei; Vineyard, Craig M. et al.
2020 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2020 - Proceedings. Institute of Electrical and Electronics Engineers Inc., 2020. p. 1598-1602 9054094 (ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings; Vol. 2020-May).

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

Dbouk, H, Geng, H, Vineyard, CM & Shanbhag, NR 2020, Low-Complexity Fixed-Point Convolutional Neural Networks for Automatic Target Recognition. in 2020 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2020 - Proceedings., 9054094, ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings, vol. 2020-May, Institute of Electrical and Electronics Engineers Inc., pp. 1598-1602, 2020 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2020, Barcelona, Spain, 5/4/20. https://doi.org/10.1109/ICASSP40776.2020.9054094

Dbouk H, Geng H, Vineyard CM, Shanbhag NR. Low-Complexity Fixed-Point Convolutional Neural Networks for Automatic Target Recognition. In 2020 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2020 - Proceedings. Institute of Electrical and Electronics Engineers Inc. 2020. p. 1598-1602. 9054094. (ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings). doi: 10.1109/ICASSP40776.2020.9054094

Dbouk, Hassan ; Geng, Hanfei ; Vineyard, Craig M. et al. / Low-Complexity Fixed-Point Convolutional Neural Networks for Automatic Target Recognition. 2020 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2020 - Proceedings. Institute of Electrical and Electronics Engineers Inc., 2020. pp. 1598-1602 (ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings).

@inproceedings{e02db1f816cc49ab9a1c6fe636653078,

title = "Low-Complexity Fixed-Point Convolutional Neural Networks for Automatic Target Recognition",

abstract = "There has been growing interest in developing neural network based automatic target recognition systems for synthetic aperture radar applications. However, these networks are typically complex in terms of storage and computation which inhibits their deployment in the field, where such resources are heavily constrained. In order to bring the cost of implementing these networks down, we develop a set of compact network architectures and train them in fixed-point. Our proposed method achieves an overall 984 reduction in terms of storage requirements and 71 × reduction in terms of computational complexity compared to state-of-the-art con-volutional neural networks for automatic target recognition (ATR), while maintaining a classification accuracy of > 99% on the MSTAR dataset.",

keywords = "automatic target recognition, deep learning, neural networks, quantization, synthetic aperture radar",

author = "Hassan Dbouk and Hanfei Geng and Vineyard, {Craig M.} and Shanbhag, {Naresh R.}",

note = "Publisher Copyright: {\textcopyright} 2020 IEEE.; 2020 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2020 ; Conference date: 04-05-2020 Through 08-05-2020",

year = "2020",

month = may,

doi = "10.1109/ICASSP40776.2020.9054094",

language = "English (US)",

series = "ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings",

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

pages = "1598--1602",

booktitle = "2020 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2020 - Proceedings",

address = "United States",

}

TY - GEN

T1 - Low-Complexity Fixed-Point Convolutional Neural Networks for Automatic Target Recognition

AU - Dbouk, Hassan

AU - Geng, Hanfei

AU - Vineyard, Craig M.

AU - Shanbhag, Naresh R.

PY - 2020/5

Y1 - 2020/5

N2 - There has been growing interest in developing neural network based automatic target recognition systems for synthetic aperture radar applications. However, these networks are typically complex in terms of storage and computation which inhibits their deployment in the field, where such resources are heavily constrained. In order to bring the cost of implementing these networks down, we develop a set of compact network architectures and train them in fixed-point. Our proposed method achieves an overall 984 reduction in terms of storage requirements and 71 × reduction in terms of computational complexity compared to state-of-the-art con-volutional neural networks for automatic target recognition (ATR), while maintaining a classification accuracy of > 99% on the MSTAR dataset.

AB - There has been growing interest in developing neural network based automatic target recognition systems for synthetic aperture radar applications. However, these networks are typically complex in terms of storage and computation which inhibits their deployment in the field, where such resources are heavily constrained. In order to bring the cost of implementing these networks down, we develop a set of compact network architectures and train them in fixed-point. Our proposed method achieves an overall 984 reduction in terms of storage requirements and 71 × reduction in terms of computational complexity compared to state-of-the-art con-volutional neural networks for automatic target recognition (ATR), while maintaining a classification accuracy of > 99% on the MSTAR dataset.

KW - automatic target recognition

KW - deep learning

KW - neural networks

KW - quantization

KW - synthetic aperture radar

UR - http://www.scopus.com/inward/record.url?scp=85089241292&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85089241292&partnerID=8YFLogxK

U2 - 10.1109/ICASSP40776.2020.9054094

DO - 10.1109/ICASSP40776.2020.9054094

M3 - Conference contribution

AN - SCOPUS:85089241292

T3 - ICASSP, IEEE International Conference on Acoustics, Speech and Signal Processing - Proceedings

SP - 1598

EP - 1602

BT - 2020 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2020 - Proceedings

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 2020 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2020

Y2 - 4 May 2020 through 8 May 2020

ER -

Low-Complexity Fixed-Point Convolutional Neural Networks for Automatic Target Recognition

Abstract

Publication series

Conference

Keywords

ASJC Scopus subject areas

Online availability

Library availability

Related links

Fingerprint

Cite this