[1]顾程成,孙 勇,程千禧,等.基于集成对比学习的高分辨遥感图像搜索模型[J].南京师范大学学报(工程技术版),2025,25(01):022-29.[doi:10.3969/j.issn.1672-1292.2025.01.003]
　Gu Chengcheng,Sun Yong,Cheng Qianxi,et al.Hi-Res Remote Image Searching Based on Ensemble Contrastive Learning[J].Journal of Nanjing Normal University(Engineering and Technology),2025,25(01):022-29.[doi:10.3969/j.issn.1672-1292.2025.01.003]
点击复制

基于集成对比学习的高分辨遥感图像搜索模型
分享到：

南京师范大学学报（工程技术版）[ISSN:1006-6977/CN:61-1281/TN]

卷:
25卷
期数:
2025年01期
页码:
022-29
栏目:
计算机科学与技术
出版日期:
2025-03-15

文章信息/Info

Title:
Hi-Res Remote Image Searching Based on Ensemble Contrastive Learning
文章编号:
1672-1292(2025)01-0022-08
作者:
顾程成1; 孙 勇1; 2; 程千禧3; 谭文安1
(1.上海第二工业大学计算机与信息工程学院,上海 201209)
(2.滁州学院实景地理环境安徽省重点实验室,安徽 滁州 239000)
(3.安徽大学资源环境学院,安徽 合肥 230009)
Author(s):
Gu Chengcheng1; Sun Yong1; 2; Cheng Qianxi3; Tan Wenan1
(1.School of Computer and Information Engineering,Shanghai Polytechnic University,Shanghai 201209,China)
(2.Key Laboratory of Physical Geographic Environment of Anhui Province,Chuzhou University,Chuzhou 239000,China)
(3.School of Resources and Environmental Engineering,Anhui University,Hefei 230009,China)
关键词:
对比学习; 集成学习; 最近邻对比学习; 动量更新对比学习; 遥感图像搜索
Keywords:
contrastive learning; ensemble learning; nearest-neighbor contrastive learning; momentum contrastive learning; remote image search
分类号:
TP751; TP18
DOI:
10.3969/j.issn.1672-1292.2025.01.003
文献标志码:
A
摘要:
为解决单一对比学习模型只关注图像的局部特征问题,提出一种基于集成对比学习的高分辨率遥感图像搜索模型. 首先,采用动量更新无监督视觉表示学习与采样最近邻对比学习模型分别提取遥感图像的局部和全局特征,以学习出更好的遥感图像视觉表征. 在此基础上,提出面向高分辨率遥感图像表征的集成对比学习模型,根据其在遥感图像分类任务的表现,对不同的对比学习器自适应地赋予不同的集成权重,通过统计对比学习特征预测准确率,进一步优化对比学习器的学习速度. 最后,将集成对比学习模型运用于高分辨遥感图像搜索. 在EuroSat、UCmerced、WHU-RS19、PatternNet等公开遥感图像数据集上的实验结果表明,所提出的模型在图像搜索任务中相较传统的对比方法有较稳定的准确率提升.
Abstract:
To address the issue of currently many single contrastive learning models focusing only on local features of images,a new high resolution remote image searching model based on ensemble contrastive learning is proposed. Firstly,local and global features and learned from remote images by momentum contrastive learning and nearest neighbor contrastive learning methods. Secondly,on this basis,high resolution remote image searching model based on ensemble contrastive learning is proposed,which sets different weights for different contrastive learners according to their own performance in downstream classification tasks,and then further optimizes the learning speeds of learners with the accuracy of contrastive learning features. Finally,ensemble contrastive learning model is applied in Hi-Res remote image searching. Experiments are conducted by using public remote sensing image datasets including EuroSat,UCmerced,WHU-RS19,and PatternNet. The results indicate that the proposed model demonstrates a more stable improvement in accuracy compared with traditional contrastive methods for image searching tasks.

参考文献/References:

[1]张重生,陈杰,李岐龙,等. 深度对比学习综述[J]. 自动化学报,2023,49(1):15-39.
[2]黄界生. 基于深度学习的计算机视觉中图像检索算法研究[J]. 信息技术与信息化,2022(9):181-184.
[3]朱鹏飞,张琬迎,王煜,等. 考虑多粒度类相关性的对比式开放集识别方法[J]. 软件学报,2022,33(4):1156-1169.
[4]DWIBEDI D,AYTAR Y,TOMPSON J,et al. With a little help from my friends:Nearest-neighbor contrastive learning of visual representations[C]//Proceedings of the 2021 IEEE/CVF International Conference on Computer Vision(ICCV). Montreal,Canada:IEEE,2021.
[5]HE K M,FAN H Q,WU Y X,et al. Momentum contrast for unsupervised visual representation learning[C]//Proceedings of the 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition(CVPR). Seattle,USA:IEEE,2020.
[6]SCHIAPPA M C,RAWAT Y S,SHAH M. Self-supervised learning for videos:A survey[J]. ACM Computing Surveys,2023,55(13s):288.
[7]GRILL J B,STRUB F,ALTCH? F,et al. Bootstrap your own latent:A new approach to self-supervised learning[J]. Advances in Neural Information Processing Systems,2020,33:21271-21284.
[8]CHEN T,KORNBLITH S,NOROUZI M,et al. A simple framework for contrastive learning of visual representations[C]//Proceedings of the 37th International Conference on Machine Learning(ICML2020). Online:ICML,2020.
[9]游文霞,申坤,杨楠,等. 基于AdaBoost集成学习的窃电检测研究[J]. 电力系统保护与控制,2020,48(19):151-159.
[10]周登文,刘子涵,刘玉铠. 基于像素对比学习的图像超分辨率算法[J]. 自动化学报,2024,50(1):181-193.
[11]LI H F,LI Y,ZHANG G,et al. Global and local contrastive self-supervised learning for semantic segmentation of HR remote sensing images[J]. IEEE Transactions on Geoscience and Remote Sensing,2022,60:5618014.
[12]ZHANG Y Z,LIU J J,SHEN W J. A review of ensemble learning algorithms used in remote sensing applications[J]. Applied Sciences,2022,12(17):8654.
[13]牟卿志,李玉婷,孙宗升,等. 集成学习算法在图像质量评估中的应用[J]. 智能计算机与应用,2023,13(10):147-150.
[14]BISWAS M,BUCKCHASH H,PRASAD D K. pNNCLR:Stochastic pseudo neighborhoods for contrastive learning based unsupervised representation learning problems[J]. Neurocomputing,2024,593:127810.
[15]LEBAILLY T,STEGM?LLER T,BOZORGTABAR B,et al. Adaptive similarity bootstrapping for self-distillation based representation learning[C]//Proceedings of the 2023 IEEE/CVF International Conference on Computer Vision(ICCV). Paris,France:IEEE,2023.
[16]GE C J,WANG J L,TONG Z,et al. Soft neighbors are positive supporters in contrastive visual representation learning[EB/OL].(2023-03-30)[2024-04-28]. https://doi.org/10.48550/arXiv.2303.17142.
[17]MIENYE I D,SUN Y X. A survey of ensemble learning:Concepts,algorithms,applications,and prospects[J]. IEEE Access,2022,10:99129-99149.
[18]FAN D,YANG D,LI X,et al. Look globally and locally:Inter-intra contrastive learning from unlabeled videos[C]//ICLR 2023 Workshop on Mathematical and Empirical Understanding of Foundation Models. Kigali,Rwanda:ICLR,2023.
[19]POWERS D M W. Evaluation:From precision,recall and F-measure to ROC,informedness,markedness & correlation[J]. Journal of Machine Learning Technology,2011,2(1):37-63.
[20]HELBER P,BISCHKE B,DENGEL A,et al. EuroSAT:A novel dataset and deep learning benchmark for land use and land cover classification[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing,2019,12(7):2217-2226.
[21]ZHOU W X,NEWSAM S,LI C M,et al. PatternNet:A benchmark dataset for performance evaluation of remote sensing image retrieval[J]. ISPRS Journal of Photogrammetry and Remote Sensing,2018,145:197-209.

备注/Memo

备注/Memo:
收稿日期:2024-04-28.
基金项目:安徽省教育厅重大重点科学研究项目(2022AH051113)、安徽省重点实验室开放基金资助项目(2022PGE003)、滁州学院科学研究基金重点项目(2022XJZD06).
通讯作者:孙勇,博士,副教授,研究方向:协同计算与地理空间人工智能. E-mail:ysun.nuaa@foxmail.com
更新日期/Last Update: 2025-03-15