{"title":"Multi-Label Classification On Aerial Images Using Deep Learning Techniques","authors":"J. Jayasree, Angaluri Venu Madhavi, G. Geetha","doi":"10.1109/ICNWC57852.2023.10127406","DOIUrl":null,"url":null,"abstract":"The one of the main problems in multi-label aerial image classification is remote sensing (RS) or aerial images understanding it increases interest in some of the research domains. Individuals can efficiently perform it by inspecting the human visual objects contained in the scene and the spatiotopological relationships of these visual objects. Although most of the existing models are pre-trained on different datasets, those existing models present some difficulties. Nowadays, Convolutional Neural Networks (CNN) have proposed a feasible approach for Aerial image Classification. With this consideration, in this work, a Deep Learning model is provided namely a convolutional neural network (CNN). In particular, CNN is employed to produce high-level appearance features and learn how visual aspects of the picture can be perceived. Our proposed models i.e., EfficeintNetB7, MobileNetV2 and ResNet50 are tested on thoroughly used datasets, and the results obtained from our proposed models show better accuracy, precision, and recall compared to the other models. Keywords - Aerial Image Classification, Convolutional Neural Network(CNN), Deep Learning, Multi-label Remote Sensing, Spatio-topological relationships.","PeriodicalId":197525,"journal":{"name":"2023 International Conference on Networking and Communications (ICNWC)","volume":null,"pages":null},"PeriodicalIF":0.0000,"publicationDate":"2023-04-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2023 International Conference on Networking and Communications (ICNWC)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ICNWC57852.2023.10127406","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
The one of the main problems in multi-label aerial image classification is remote sensing (RS) or aerial images understanding it increases interest in some of the research domains. Individuals can efficiently perform it by inspecting the human visual objects contained in the scene and the spatiotopological relationships of these visual objects. Although most of the existing models are pre-trained on different datasets, those existing models present some difficulties. Nowadays, Convolutional Neural Networks (CNN) have proposed a feasible approach for Aerial image Classification. With this consideration, in this work, a Deep Learning model is provided namely a convolutional neural network (CNN). In particular, CNN is employed to produce high-level appearance features and learn how visual aspects of the picture can be perceived. Our proposed models i.e., EfficeintNetB7, MobileNetV2 and ResNet50 are tested on thoroughly used datasets, and the results obtained from our proposed models show better accuracy, precision, and recall compared to the other models. Keywords - Aerial Image Classification, Convolutional Neural Network(CNN), Deep Learning, Multi-label Remote Sensing, Spatio-topological relationships.