计算机工程与应用2024，Vol.60Issue(4)：173-182,10.DOI:10.3778/j.issn.1002-8331.2301-0012
融入注意力机制的小样本遥感图像场景分类
Few-Shot Scene Classification with Attention Mechanism in Remote Sensing
摘要
Abstract
Remote sensing scene classification is a hot research topic in the field of computer vision,and it is of great sig-nificance to semantic understanding of remote sensing images.At present,remote sensing scene classification methods based on deep learning occupy a dominant position in this field.However,it suffers from the lack of samples and poor model generalization ability in actual application scenarios.Therefore,this paper proposes a few-shot remote scene classi-fication method based on attention mechanism,and designs a structure of dual-branches similarity measurement.This method is based on the meta-learning training strategy to divide the dataset into tasks.At the meantime,the input images are divided into blocks in order to preserve the feature distribution in the remote sensing image.Then the lightweight at-tention module is introduced into the feature extraction network to reduce the risk of overfitting and ensure the acquisition of discriminative features.Finally,based on earth mover's distance(EMD),a dual-branches similarity measurement mod-ule is added to improve the discriminative ability of the classifier.The results show that compared with the classic small-sample learning method,the few-shot remote scene classification method proposed in this paper can significantly improve the classification performance.关键词
遥感图像场景分类/小样本学习/元学习/注意力机制/双分支判别Key words
remote sensing scene classification/few-shot learning/meta-learning/attention mechanism/dual-branches similarity measurement分类
信息技术与安全科学引用本文复制引用
张多纳,赵宏佳,鲁远耀,崔健,张宝昌..融入注意力机制的小样本遥感图像场景分类[J].计算机工程与应用,2024,60(4):173-182,10.基金项目
国家自然科学基金(62201010,U20B2042,61971007,61571013) （62201010,U20B2042,61971007,61571013）
北京市教育委员会科研计划项目(KM202310009003) （KM202310009003）
北方工业大学科研启动基金. （）
