Self-Supervised GANs With Similarity Loss for Remote Sensing Image Scene Classification
With the development of supervised deep neural networks, classification performance on existing remote sensing scene datasets has been markedly improved. However, supervised learning methods rely heavily on large-scale tagged examples to obtain a better prediction performance. The lack of large-scal...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2021-01-01
|
| Series: | IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/9345971/ |
| _version_ | 1818890095788294144 |
|---|---|
| author | Dongen Guo Ying Xia Xiaobo Luo |
| author_facet | Dongen Guo Ying Xia Xiaobo Luo |
| author_sort | Dongen Guo |
| collection | DOAJ |
| description | With the development of supervised deep neural networks, classification performance on existing remote sensing scene datasets has been markedly improved. However, supervised learning methods rely heavily on large-scale tagged examples to obtain a better prediction performance. The lack of large-scale tagged remote sensing scene images has become the primary bottleneck in scene classification. To deal with this issue, a novel scene classification method using self-supervised gated self-attention generative adversarial networks (GANs) with similarity loss is proposed. Specifically, the gated self-attention module is first introduced into GANs to focus on key scene areas and filter useless information for strengthening feature representations. Then, the pyramidal convolution block is introduced into the residual block of the discriminator to capture different levels of details in the image using different types of filters with varying sizes and depths for enhancing the feature representations of the discriminator. Additionally, a novel similarity loss item is integrated into the discriminator to leverage self-supervised learning. Besides, spectral normalization is introduced into both the generative network and discriminative network to stabilize training and enhance feature representations. The architecture of multilevel feature fusion is integrated into the discriminative network to achieve more discriminant representation. Experimental results on the AID and NWPU-RESISC45 datasets show that the proposed method achieves the best performance compared to the existing unsupervised classification methods. |
| first_indexed | 2024-12-19T17:19:28Z |
| format | Article |
| id | doaj.art-eb7142200a5f43abbf7a6e806a600ab0 |
| institution | Directory Open Access Journal |
| issn | 2151-1535 |
| language | English |
| last_indexed | 2024-12-19T17:19:28Z |
| publishDate | 2021-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
| spelling | doaj.art-eb7142200a5f43abbf7a6e806a600ab02022-12-21T20:12:43ZengIEEEIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing2151-15352021-01-01142508252110.1109/JSTARS.2021.30568839345971Self-Supervised GANs With Similarity Loss for Remote Sensing Image Scene ClassificationDongen Guo0https://orcid.org/0000-0003-3927-7616Ying Xia1https://orcid.org/0000-0002-7407-6126Xiaobo Luo2https://orcid.org/0000-0001-5688-0324Chongqing Engineering Research Center for Spatial Big Data Intelligent Technology, Chongqing University of Posts and Telecommunications, Chongqing, ChinaChongqing Engineering Research Center for Spatial Big Data Intelligent Technology, Chongqing University of Posts and Telecommunications, Chongqing, ChinaChongqing Engineering Research Center for Spatial Big Data Intelligent Technology, Chongqing University of Posts and Telecommunications, Chongqing, ChinaWith the development of supervised deep neural networks, classification performance on existing remote sensing scene datasets has been markedly improved. However, supervised learning methods rely heavily on large-scale tagged examples to obtain a better prediction performance. The lack of large-scale tagged remote sensing scene images has become the primary bottleneck in scene classification. To deal with this issue, a novel scene classification method using self-supervised gated self-attention generative adversarial networks (GANs) with similarity loss is proposed. Specifically, the gated self-attention module is first introduced into GANs to focus on key scene areas and filter useless information for strengthening feature representations. Then, the pyramidal convolution block is introduced into the residual block of the discriminator to capture different levels of details in the image using different types of filters with varying sizes and depths for enhancing the feature representations of the discriminator. Additionally, a novel similarity loss item is integrated into the discriminator to leverage self-supervised learning. Besides, spectral normalization is introduced into both the generative network and discriminative network to stabilize training and enhance feature representations. The architecture of multilevel feature fusion is integrated into the discriminative network to achieve more discriminant representation. Experimental results on the AID and NWPU-RESISC45 datasets show that the proposed method achieves the best performance compared to the existing unsupervised classification methods.https://ieeexplore.ieee.org/document/9345971/Gated self-attention (GAS) modulegenerative adversarial networks (GANs)pyramidal convolution (PyConv)remote sensing image scene classificationself-supervised learningsimilarity loss |
| spellingShingle | Dongen Guo Ying Xia Xiaobo Luo Self-Supervised GANs With Similarity Loss for Remote Sensing Image Scene Classification IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing Gated self-attention (GAS) module generative adversarial networks (GANs) pyramidal convolution (PyConv) remote sensing image scene classification self-supervised learning similarity loss |
| title | Self-Supervised GANs With Similarity Loss for Remote Sensing Image Scene Classification |
| title_full | Self-Supervised GANs With Similarity Loss for Remote Sensing Image Scene Classification |
| title_fullStr | Self-Supervised GANs With Similarity Loss for Remote Sensing Image Scene Classification |
| title_full_unstemmed | Self-Supervised GANs With Similarity Loss for Remote Sensing Image Scene Classification |
| title_short | Self-Supervised GANs With Similarity Loss for Remote Sensing Image Scene Classification |
| title_sort | self supervised gans with similarity loss for remote sensing image scene classification |
| topic | Gated self-attention (GAS) module generative adversarial networks (GANs) pyramidal convolution (PyConv) remote sensing image scene classification self-supervised learning similarity loss |
| url | https://ieeexplore.ieee.org/document/9345971/ |
| work_keys_str_mv | AT dongenguo selfsupervisedganswithsimilaritylossforremotesensingimagesceneclassification AT yingxia selfsupervisedganswithsimilaritylossforremotesensingimagesceneclassification AT xiaoboluo selfsupervisedganswithsimilaritylossforremotesensingimagesceneclassification |