DML-GANR: Deep Metric Learning With Generative Adversarial Network Regularization for High Spatial Resolution Remote Sensing Image Retrieval
With a small number of labeled samples for training, it can save considerable manpower and material resources, especially when the amount of high spatial resolution remote sensing images (HSR-RSIs) increases considerably. However, many deep models face the problem of overfitting when using a small n...
Main Authors: | , , , , , , |
---|---|
Format: | Article in Journal/Newspaper |
Language: | unknown |
Published: |
arXiv
2020
|
Subjects: | |
Online Access: | https://dx.doi.org/10.48550/arxiv.2010.03116 https://arxiv.org/abs/2010.03116 |
id |
ftdatacite:10.48550/arxiv.2010.03116 |
---|---|
record_format |
openpolar |
spelling |
ftdatacite:10.48550/arxiv.2010.03116 2023-05-15T16:01:11+02:00 DML-GANR: Deep Metric Learning With Generative Adversarial Network Regularization for High Spatial Resolution Remote Sensing Image Retrieval Cao, Yun Wang, Yuebin Peng, Junhuan Zhang, Liqiang Xu, Linlin Yan, Kai Li, Lihua 2020 https://dx.doi.org/10.48550/arxiv.2010.03116 https://arxiv.org/abs/2010.03116 unknown arXiv https://dx.doi.org/10.1109/tgrs.2020.2991545 arXiv.org perpetual, non-exclusive license http://arxiv.org/licenses/nonexclusive-distrib/1.0/ Computer Vision and Pattern Recognition cs.CV Optimization and Control math.OC FOS Computer and information sciences FOS Mathematics article-journal Article ScholarlyArticle Text 2020 ftdatacite https://doi.org/10.48550/arxiv.2010.03116 https://doi.org/10.1109/tgrs.2020.2991545 2022-03-10T15:01:56Z With a small number of labeled samples for training, it can save considerable manpower and material resources, especially when the amount of high spatial resolution remote sensing images (HSR-RSIs) increases considerably. However, many deep models face the problem of overfitting when using a small number of labeled samples. This might degrade HSRRSI retrieval accuracy. Aiming at obtaining more accurate HSR-RSI retrieval performance with small training samples, we develop a deep metric learning approach with generative adversarial network regularization (DML-GANR) for HSR-RSI retrieval. The DML-GANR starts from a high-level feature extraction (HFE) to extract high-level features, which includes convolutional layers and fully connected (FC) layers. Each of the FC layers is constructed by deep metric learning (DML) to maximize the interclass variations and minimize the intraclass variations. The generative adversarial network (GAN) is adopted to mitigate the overfitting problem and validate the qualities of extracted high-level features. DML-GANR is optimized through a customized approach, and the optimal parameters are obtained. The experimental results on the three data sets demonstrate the superior performance of DML-GANR over state-of-the-art techniques in HSR-RSI retrieval. : 17 pages Article in Journal/Newspaper DML DataCite Metadata Store (German National Library of Science and Technology) |
institution |
Open Polar |
collection |
DataCite Metadata Store (German National Library of Science and Technology) |
op_collection_id |
ftdatacite |
language |
unknown |
topic |
Computer Vision and Pattern Recognition cs.CV Optimization and Control math.OC FOS Computer and information sciences FOS Mathematics |
spellingShingle |
Computer Vision and Pattern Recognition cs.CV Optimization and Control math.OC FOS Computer and information sciences FOS Mathematics Cao, Yun Wang, Yuebin Peng, Junhuan Zhang, Liqiang Xu, Linlin Yan, Kai Li, Lihua DML-GANR: Deep Metric Learning With Generative Adversarial Network Regularization for High Spatial Resolution Remote Sensing Image Retrieval |
topic_facet |
Computer Vision and Pattern Recognition cs.CV Optimization and Control math.OC FOS Computer and information sciences FOS Mathematics |
description |
With a small number of labeled samples for training, it can save considerable manpower and material resources, especially when the amount of high spatial resolution remote sensing images (HSR-RSIs) increases considerably. However, many deep models face the problem of overfitting when using a small number of labeled samples. This might degrade HSRRSI retrieval accuracy. Aiming at obtaining more accurate HSR-RSI retrieval performance with small training samples, we develop a deep metric learning approach with generative adversarial network regularization (DML-GANR) for HSR-RSI retrieval. The DML-GANR starts from a high-level feature extraction (HFE) to extract high-level features, which includes convolutional layers and fully connected (FC) layers. Each of the FC layers is constructed by deep metric learning (DML) to maximize the interclass variations and minimize the intraclass variations. The generative adversarial network (GAN) is adopted to mitigate the overfitting problem and validate the qualities of extracted high-level features. DML-GANR is optimized through a customized approach, and the optimal parameters are obtained. The experimental results on the three data sets demonstrate the superior performance of DML-GANR over state-of-the-art techniques in HSR-RSI retrieval. : 17 pages |
format |
Article in Journal/Newspaper |
author |
Cao, Yun Wang, Yuebin Peng, Junhuan Zhang, Liqiang Xu, Linlin Yan, Kai Li, Lihua |
author_facet |
Cao, Yun Wang, Yuebin Peng, Junhuan Zhang, Liqiang Xu, Linlin Yan, Kai Li, Lihua |
author_sort |
Cao, Yun |
title |
DML-GANR: Deep Metric Learning With Generative Adversarial Network Regularization for High Spatial Resolution Remote Sensing Image Retrieval |
title_short |
DML-GANR: Deep Metric Learning With Generative Adversarial Network Regularization for High Spatial Resolution Remote Sensing Image Retrieval |
title_full |
DML-GANR: Deep Metric Learning With Generative Adversarial Network Regularization for High Spatial Resolution Remote Sensing Image Retrieval |
title_fullStr |
DML-GANR: Deep Metric Learning With Generative Adversarial Network Regularization for High Spatial Resolution Remote Sensing Image Retrieval |
title_full_unstemmed |
DML-GANR: Deep Metric Learning With Generative Adversarial Network Regularization for High Spatial Resolution Remote Sensing Image Retrieval |
title_sort |
dml-ganr: deep metric learning with generative adversarial network regularization for high spatial resolution remote sensing image retrieval |
publisher |
arXiv |
publishDate |
2020 |
url |
https://dx.doi.org/10.48550/arxiv.2010.03116 https://arxiv.org/abs/2010.03116 |
genre |
DML |
genre_facet |
DML |
op_relation |
https://dx.doi.org/10.1109/tgrs.2020.2991545 |
op_rights |
arXiv.org perpetual, non-exclusive license http://arxiv.org/licenses/nonexclusive-distrib/1.0/ |
op_doi |
https://doi.org/10.48550/arxiv.2010.03116 https://doi.org/10.1109/tgrs.2020.2991545 |
_version_ |
1766397153270300672 |