A Triplet Nonlocal Neural Network With Dual-Anchor Triplet Loss for High-Resolution Remote Sensing Image Retrieval
Conventional deep-learning-based retrieval models are generally trained under the framework of scene classification with cross-entropy loss, this way focuses only on the output probability corresponding to the label of input samples, while ignoring the predictive information of other categories, whi...
Published in: | IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
---|---|
Main Authors: | , , , |
Format: | Article in Journal/Newspaper |
Language: | English |
Published: |
IEEE
2021
|
Subjects: | |
Online Access: | https://doi.org/10.1109/JSTARS.2021.3058691 https://doaj.org/article/bbdf70622b0249e8a392a19122d2d4e1 |
id |
ftdoajarticles:oai:doaj.org/article:bbdf70622b0249e8a392a19122d2d4e1 |
---|---|
record_format |
openpolar |
spelling |
ftdoajarticles:oai:doaj.org/article:bbdf70622b0249e8a392a19122d2d4e1 2023-05-15T16:02:02+02:00 A Triplet Nonlocal Neural Network With Dual-Anchor Triplet Loss for High-Resolution Remote Sensing Image Retrieval Maoding Zhang Qimin Cheng Fang Luo Lan Ye 2021-01-01T00:00:00Z https://doi.org/10.1109/JSTARS.2021.3058691 https://doaj.org/article/bbdf70622b0249e8a392a19122d2d4e1 EN eng IEEE https://ieeexplore.ieee.org/document/9353191/ https://doaj.org/toc/2151-1535 2151-1535 doi:10.1109/JSTARS.2021.3058691 https://doaj.org/article/bbdf70622b0249e8a392a19122d2d4e1 IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, Vol 14, Pp 2711-2723 (2021) Deep metric learning (DML) dual-anchor triplet loss high-resolution remote sensing image (HRRSI) retrieval triplet nonlocal neural network (T-NLNN) Ocean engineering TC1501-1800 Geophysics. Cosmic physics QC801-809 article 2021 ftdoajarticles https://doi.org/10.1109/JSTARS.2021.3058691 2022-12-31T09:11:47Z Conventional deep-learning-based retrieval models are generally trained under the framework of scene classification with cross-entropy loss, this way focuses only on the output probability corresponding to the label of input samples, while ignoring the predictive information of other categories, which makes the retrieval accuracy susceptible to the intraclass difference of the image samples. And conventional methods often used fixed-size convolution kernels that only consider the local area with fixed sizes, thus largely ignoring the global information. In response to the above problems, this article constructs a triplet nonlocal neural network (T-NLNN) model that combines deep metric learning and nonlocal operation. The proposed T-NLNN follows the three-branch network design, with shared weights in each branch. We evaluate T-NLNN on three public high-resolution remote sensing datasets, and the experimental results suggest that T-NLNN has discriminative feature learning ability and outperforms other existing algorithms. In addition, we propose a dual-anchor triplet loss function to facilitate the utilization of information in the input samples. The experimental results prove that the proposed dual-anchor triplet loss function works better than the traditional triplet loss function on all datasets. Article in Journal/Newspaper DML Directory of Open Access Journals: DOAJ Articles IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing 14 2711 2723 |
institution |
Open Polar |
collection |
Directory of Open Access Journals: DOAJ Articles |
op_collection_id |
ftdoajarticles |
language |
English |
topic |
Deep metric learning (DML) dual-anchor triplet loss high-resolution remote sensing image (HRRSI) retrieval triplet nonlocal neural network (T-NLNN) Ocean engineering TC1501-1800 Geophysics. Cosmic physics QC801-809 |
spellingShingle |
Deep metric learning (DML) dual-anchor triplet loss high-resolution remote sensing image (HRRSI) retrieval triplet nonlocal neural network (T-NLNN) Ocean engineering TC1501-1800 Geophysics. Cosmic physics QC801-809 Maoding Zhang Qimin Cheng Fang Luo Lan Ye A Triplet Nonlocal Neural Network With Dual-Anchor Triplet Loss for High-Resolution Remote Sensing Image Retrieval |
topic_facet |
Deep metric learning (DML) dual-anchor triplet loss high-resolution remote sensing image (HRRSI) retrieval triplet nonlocal neural network (T-NLNN) Ocean engineering TC1501-1800 Geophysics. Cosmic physics QC801-809 |
description |
Conventional deep-learning-based retrieval models are generally trained under the framework of scene classification with cross-entropy loss, this way focuses only on the output probability corresponding to the label of input samples, while ignoring the predictive information of other categories, which makes the retrieval accuracy susceptible to the intraclass difference of the image samples. And conventional methods often used fixed-size convolution kernels that only consider the local area with fixed sizes, thus largely ignoring the global information. In response to the above problems, this article constructs a triplet nonlocal neural network (T-NLNN) model that combines deep metric learning and nonlocal operation. The proposed T-NLNN follows the three-branch network design, with shared weights in each branch. We evaluate T-NLNN on three public high-resolution remote sensing datasets, and the experimental results suggest that T-NLNN has discriminative feature learning ability and outperforms other existing algorithms. In addition, we propose a dual-anchor triplet loss function to facilitate the utilization of information in the input samples. The experimental results prove that the proposed dual-anchor triplet loss function works better than the traditional triplet loss function on all datasets. |
format |
Article in Journal/Newspaper |
author |
Maoding Zhang Qimin Cheng Fang Luo Lan Ye |
author_facet |
Maoding Zhang Qimin Cheng Fang Luo Lan Ye |
author_sort |
Maoding Zhang |
title |
A Triplet Nonlocal Neural Network With Dual-Anchor Triplet Loss for High-Resolution Remote Sensing Image Retrieval |
title_short |
A Triplet Nonlocal Neural Network With Dual-Anchor Triplet Loss for High-Resolution Remote Sensing Image Retrieval |
title_full |
A Triplet Nonlocal Neural Network With Dual-Anchor Triplet Loss for High-Resolution Remote Sensing Image Retrieval |
title_fullStr |
A Triplet Nonlocal Neural Network With Dual-Anchor Triplet Loss for High-Resolution Remote Sensing Image Retrieval |
title_full_unstemmed |
A Triplet Nonlocal Neural Network With Dual-Anchor Triplet Loss for High-Resolution Remote Sensing Image Retrieval |
title_sort |
triplet nonlocal neural network with dual-anchor triplet loss for high-resolution remote sensing image retrieval |
publisher |
IEEE |
publishDate |
2021 |
url |
https://doi.org/10.1109/JSTARS.2021.3058691 https://doaj.org/article/bbdf70622b0249e8a392a19122d2d4e1 |
genre |
DML |
genre_facet |
DML |
op_source |
IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, Vol 14, Pp 2711-2723 (2021) |
op_relation |
https://ieeexplore.ieee.org/document/9353191/ https://doaj.org/toc/2151-1535 2151-1535 doi:10.1109/JSTARS.2021.3058691 https://doaj.org/article/bbdf70622b0249e8a392a19122d2d4e1 |
op_doi |
https://doi.org/10.1109/JSTARS.2021.3058691 |
container_title |
IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing |
container_volume |
14 |
container_start_page |
2711 |
op_container_end_page |
2723 |
_version_ |
1766397666869116928 |