Change Detection Network Based on Transformer and Transfer Learning

The purpose of the change detection(CD) task is to contrast the change information of a specific object in remote sensing images from different time periods. The deep-learning-based change-detection algorithm can extract pixel-level semantic segmentation results for changed objects. Currently, deep...

Full description

Saved in:
Bibliographic Details
Main Authors: Hua Li, Jingyu Li, Guanghao Luo, Liang Zhou, Hao Wu, Zhangcai Yin
Format: Article
Language:English
Published: IEEE 2025-01-01
Series:IEEE Access
Subjects:
Online Access:https://ieeexplore.ieee.org/document/11020642/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850100760567611392
author Hua Li
Jingyu Li
Guanghao Luo
Liang Zhou
Hao Wu
Zhangcai Yin
author_facet Hua Li
Jingyu Li
Guanghao Luo
Liang Zhou
Hao Wu
Zhangcai Yin
author_sort Hua Li
collection DOAJ
description The purpose of the change detection(CD) task is to contrast the change information of a specific object in remote sensing images from different time periods. The deep-learning-based change-detection algorithm can extract pixel-level semantic segmentation results for changed objects. Currently, deep learning based change detection algorithms have achieved excellent detection results through the meticulous design of feature extraction and change judgment modules. However, challenges, such as limited generalization ability owing to small-scale change-detection datasets and noise generated by the change judgment module, still exist. Therefore, it is crucial to enhance the generalization capability of the change detection model and improve its ability to detect challenging information in remote sensing building change tasks. In this study, we propose a swin transformer based change detection network (STTL-CD) that incorporates Transfer Learning. Based on the idea of transfer learning, we transferred the weights of the backbone network trained on a large semantic segmentation dataset to initialize our feature extraction network, and enhance the generalization ability of the change detection network. To address the issue of noise during change detection, we established a joint loss function that increases the weight ratio of difficult and easy samples, thereby compelling the model to focus on discerning challenging samples. Our experimental results demonstrate that STTL-CD achieves IoU(0.8571), F1(0.9231) and IoU(0.8902), F1(0.9422) scores in the LEVIR-CD and WHU-CD datasets, surpassing the state-of-the-art change-detection methods developed in recent years.
format Article
id doaj-art-5be74e46f24e45a485d28cc04dab2fed
institution DOAJ
issn 2169-3536
language English
publishDate 2025-01-01
publisher IEEE
record_format Article
series IEEE Access
spelling doaj-art-5be74e46f24e45a485d28cc04dab2fed2025-08-20T02:40:13ZengIEEEIEEE Access2169-35362025-01-011310113110114210.1109/ACCESS.2025.357548311020642Change Detection Network Based on Transformer and Transfer LearningHua Li0https://orcid.org/0000-0002-7043-6374Jingyu Li1https://orcid.org/0009-0003-1603-1207Guanghao Luo2https://orcid.org/0009-0009-4046-3484Liang Zhou3Hao Wu4https://orcid.org/0000-0001-5751-7885Zhangcai Yin5School of Resource and Environmental Engineering, Wuhan University of Technology, Wuhan, ChinaSchool of Resource and Environmental Engineering, Wuhan University of Technology, Wuhan, ChinaSchool of Resource and Environmental Engineering, Wuhan University of Technology, Wuhan, ChinaSchool of Resource and Environmental Engineering, Wuhan University of Technology, Wuhan, ChinaCollege of Urban and Environment Science, Central China Normal University, Wuhan, ChinaSchool of Resource and Environmental Engineering, Wuhan University of Technology, Wuhan, ChinaThe purpose of the change detection(CD) task is to contrast the change information of a specific object in remote sensing images from different time periods. The deep-learning-based change-detection algorithm can extract pixel-level semantic segmentation results for changed objects. Currently, deep learning based change detection algorithms have achieved excellent detection results through the meticulous design of feature extraction and change judgment modules. However, challenges, such as limited generalization ability owing to small-scale change-detection datasets and noise generated by the change judgment module, still exist. Therefore, it is crucial to enhance the generalization capability of the change detection model and improve its ability to detect challenging information in remote sensing building change tasks. In this study, we propose a swin transformer based change detection network (STTL-CD) that incorporates Transfer Learning. Based on the idea of transfer learning, we transferred the weights of the backbone network trained on a large semantic segmentation dataset to initialize our feature extraction network, and enhance the generalization ability of the change detection network. To address the issue of noise during change detection, we established a joint loss function that increases the weight ratio of difficult and easy samples, thereby compelling the model to focus on discerning challenging samples. Our experimental results demonstrate that STTL-CD achieves IoU(0.8571), F1(0.9231) and IoU(0.8902), F1(0.9422) scores in the LEVIR-CD and WHU-CD datasets, surpassing the state-of-the-art change-detection methods developed in recent years.https://ieeexplore.ieee.org/document/11020642/Change detectionfocal lossremote sensing imagestransfer learningtransformer
spellingShingle Hua Li
Jingyu Li
Guanghao Luo
Liang Zhou
Hao Wu
Zhangcai Yin
Change Detection Network Based on Transformer and Transfer Learning
IEEE Access
Change detection
focal loss
remote sensing images
transfer learning
transformer
title Change Detection Network Based on Transformer and Transfer Learning
title_full Change Detection Network Based on Transformer and Transfer Learning
title_fullStr Change Detection Network Based on Transformer and Transfer Learning
title_full_unstemmed Change Detection Network Based on Transformer and Transfer Learning
title_short Change Detection Network Based on Transformer and Transfer Learning
title_sort change detection network based on transformer and transfer learning
topic Change detection
focal loss
remote sensing images
transfer learning
transformer
url https://ieeexplore.ieee.org/document/11020642/
work_keys_str_mv AT huali changedetectionnetworkbasedontransformerandtransferlearning
AT jingyuli changedetectionnetworkbasedontransformerandtransferlearning
AT guanghaoluo changedetectionnetworkbasedontransformerandtransferlearning
AT liangzhou changedetectionnetworkbasedontransformerandtransferlearning
AT haowu changedetectionnetworkbasedontransformerandtransferlearning
AT zhangcaiyin changedetectionnetworkbasedontransformerandtransferlearning