SAR Incremental Learning via Generative Adversarial Network and Experience Replay

Fei Gao; Chen Fan; Penghui Chen; Amir Hussain; Huiyu Zhou

doi:10.1007/978-981-97-1417-9_32

SAR Incremental Learning via Generative Adversarial Network and Experience Replay

Fei Gao
, Chen Fan
, Penghui Chen^*
, Amir Hussain
, Huiyu Zhou

^*Corresponding author for this work

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

Abstract

Incremental learning is a viable approach for addressing the recognition problem of SAR images in data stream scenarios. However, it may suffer from catastrophic forgetting due to insufficient exposure to old category data during training. This paper proposes the integration of a generative adversarial network (GAN) into the iCaRL model to mitigate catastrophic forgetting by generating samples from previously learned categories. Experimental results on our SAR dataset demonstrate that our approach significantly enhances the recognition performance of the iCaRL model.

Original language	English
Title of host publication	Advances in Brain Inspired Cognitive Systems - 13th International Conference, BICS 2023, Proceedings
Editors	Jinchang Ren, Amir Hussain, Iman Yi Liao, Rongjun Chen, Kaizhu Huang, Huimin Zhao, Xiaoyong Liu, Ping Ma, Thomas Maul
Publisher	Springer Science and Business Media Deutschland GmbH
Pages	345-353
Number of pages	9
ISBN (Print)	9789819714162
DOIs	https://doi.org/10.1007/978-981-97-1417-9_32
State	Published - 2024
Externally published	Yes
Event	13th International Conference on Brain Inspired Cognitive Systems, BICS 2023 - Kuala Lumpur, Malaysia Duration: 5 Aug 2023 → 6 Aug 2023

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	14374 LNAI
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	13th International Conference on Brain Inspired Cognitive Systems, BICS 2023
Country/Territory	Malaysia
City	Kuala Lumpur
Period	5/08/23 → 6/08/23

Bibliographical note

Publisher Copyright:
© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024.

Keywords

Generative adversarial network
Incremental learning
SAR target recognition

ASJC Scopus subject areas

Theoretical Computer Science
General Computer Science

Access to Document

10.1007/978-981-97-1417-9_32

Cite this

Gao, F., Fan, C., Chen, P., Hussain, A., & Zhou, H. (2024). SAR Incremental Learning via Generative Adversarial Network and Experience Replay. In J. Ren, A. Hussain, I. Y. Liao, R. Chen, K. Huang, H. Zhao, X. Liu, P. Ma, & T. Maul (Eds.), Advances in Brain Inspired Cognitive Systems - 13th International Conference, BICS 2023, Proceedings (pp. 345-353). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 14374 LNAI). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-981-97-1417-9_32

Gao, Fei ; Fan, Chen ; Chen, Penghui et al. / SAR Incremental Learning via Generative Adversarial Network and Experience Replay. Advances in Brain Inspired Cognitive Systems - 13th International Conference, BICS 2023, Proceedings. editor / Jinchang Ren ; Amir Hussain ; Iman Yi Liao ; Rongjun Chen ; Kaizhu Huang ; Huimin Zhao ; Xiaoyong Liu ; Ping Ma ; Thomas Maul. Springer Science and Business Media Deutschland GmbH, 2024. pp. 345-353 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{8f47604675b14f23b8d89a5eedb57803,

title = "SAR Incremental Learning via Generative Adversarial Network and Experience Replay",

abstract = "Incremental learning is a viable approach for addressing the recognition problem of SAR images in data stream scenarios. However, it may suffer from catastrophic forgetting due to insufficient exposure to old category data during training. This paper proposes the integration of a generative adversarial network (GAN) into the iCaRL model to mitigate catastrophic forgetting by generating samples from previously learned categories. Experimental results on our SAR dataset demonstrate that our approach significantly enhances the recognition performance of the iCaRL model.",

keywords = "Generative adversarial network, Incremental learning, SAR target recognition",

author = "Fei Gao and Chen Fan and Penghui Chen and Amir Hussain and Huiyu Zhou",

note = "Publisher Copyright: {\textcopyright} The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024.; 13th International Conference on Brain Inspired Cognitive Systems, BICS 2023 ; Conference date: 05-08-2023 Through 06-08-2023",

year = "2024",

doi = "10.1007/978-981-97-1417-9\_32",

language = "English",

isbn = "9789819714162",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Science and Business Media Deutschland GmbH",

pages = "345--353",

editor = "Jinchang Ren and Amir Hussain and Liao, \{Iman Yi\} and Rongjun Chen and Kaizhu Huang and Huimin Zhao and Xiaoyong Liu and Ping Ma and Thomas Maul",

booktitle = "Advances in Brain Inspired Cognitive Systems - 13th International Conference, BICS 2023, Proceedings",

address = "Germany",

}

Gao, F, Fan, C, Chen, P, Hussain, A & Zhou, H 2024, SAR Incremental Learning via Generative Adversarial Network and Experience Replay. in J Ren, A Hussain, IY Liao, R Chen, K Huang, H Zhao, X Liu, P Ma & T Maul (eds), Advances in Brain Inspired Cognitive Systems - 13th International Conference, BICS 2023, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 14374 LNAI, Springer Science and Business Media Deutschland GmbH, pp. 345-353, 13th International Conference on Brain Inspired Cognitive Systems, BICS 2023, Kuala Lumpur, Malaysia, 5/08/23. https://doi.org/10.1007/978-981-97-1417-9_32

SAR Incremental Learning via Generative Adversarial Network and Experience Replay. / Gao, Fei; Fan, Chen; Chen, Penghui et al.
Advances in Brain Inspired Cognitive Systems - 13th International Conference, BICS 2023, Proceedings. ed. / Jinchang Ren; Amir Hussain; Iman Yi Liao; Rongjun Chen; Kaizhu Huang; Huimin Zhao; Xiaoyong Liu; Ping Ma; Thomas Maul. Springer Science and Business Media Deutschland GmbH, 2024. p. 345-353 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 14374 LNAI).

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

TY - GEN

T1 - SAR Incremental Learning via Generative Adversarial Network and Experience Replay

AU - Gao, Fei

AU - Fan, Chen

AU - Chen, Penghui

AU - Hussain, Amir

AU - Zhou, Huiyu

N1 - Publisher Copyright: © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024.

PY - 2024

Y1 - 2024

N2 - Incremental learning is a viable approach for addressing the recognition problem of SAR images in data stream scenarios. However, it may suffer from catastrophic forgetting due to insufficient exposure to old category data during training. This paper proposes the integration of a generative adversarial network (GAN) into the iCaRL model to mitigate catastrophic forgetting by generating samples from previously learned categories. Experimental results on our SAR dataset demonstrate that our approach significantly enhances the recognition performance of the iCaRL model.

AB - Incremental learning is a viable approach for addressing the recognition problem of SAR images in data stream scenarios. However, it may suffer from catastrophic forgetting due to insufficient exposure to old category data during training. This paper proposes the integration of a generative adversarial network (GAN) into the iCaRL model to mitigate catastrophic forgetting by generating samples from previously learned categories. Experimental results on our SAR dataset demonstrate that our approach significantly enhances the recognition performance of the iCaRL model.

KW - Generative adversarial network

KW - Incremental learning

KW - SAR target recognition

UR - https://www.scopus.com/pages/publications/85195124220

U2 - 10.1007/978-981-97-1417-9_32

DO - 10.1007/978-981-97-1417-9_32

M3 - Conference contribution

AN - SCOPUS:85195124220

SN - 9789819714162

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 345

EP - 353

BT - Advances in Brain Inspired Cognitive Systems - 13th International Conference, BICS 2023, Proceedings

A2 - Ren, Jinchang

A2 - Hussain, Amir

A2 - Liao, Iman Yi

A2 - Chen, Rongjun

A2 - Huang, Kaizhu

A2 - Zhao, Huimin

A2 - Liu, Xiaoyong

A2 - Ma, Ping

A2 - Maul, Thomas

PB - Springer Science and Business Media Deutschland GmbH

T2 - 13th International Conference on Brain Inspired Cognitive Systems, BICS 2023

Y2 - 5 August 2023 through 6 August 2023

ER -

Gao F, Fan C, Chen P, Hussain A, Zhou H. SAR Incremental Learning via Generative Adversarial Network and Experience Replay. In Ren J, Hussain A, Liao IY, Chen R, Huang K, Zhao H, Liu X, Ma P, Maul T, editors, Advances in Brain Inspired Cognitive Systems - 13th International Conference, BICS 2023, Proceedings. Springer Science and Business Media Deutschland GmbH. 2024. p. 345-353. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-981-97-1417-9_32

SAR Incremental Learning via Generative Adversarial Network and Experience Replay

Abstract

Publication series

Conference

Bibliographical note

Keywords

ASJC Scopus subject areas

Access to Document

Fingerprint

Cite this