{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/deepsat-a-learning-framework-for-satellite","title":"DeepSat - A Learning framework for Satellite Imagery","arxiv_id":"1509.03602","date":"2015-09-11","proceeding":null,"authors":["Saikat Basu","Sangram Ganguly","Supratik Mukhopadhyay","Robert DiBiano","Manohar Karki","Ramakrishna Nemani"],"abstract":"Satellite image classification is a challenging problem that lies at the\ncrossroads of remote sensing, computer vision, and machine learning. Due to the\nhigh variability inherent in satellite data, most of the current object\nclassification approaches are not suitable for handling satellite datasets. The\nprogress of satellite image analytics has also been inhibited by the lack of a\nsingle labeled high-resolution dataset with multiple class labels. The\ncontributions of this paper are twofold - (1) first, we present two new\nsatellite datasets called SAT-4 and SAT-6, and (2) then, we propose a\nclassification framework that extracts features from an input image, normalizes\nthem and feeds the normalized feature vectors to a Deep Belief Network for\nclassification. On the SAT-4 dataset, our best network produces a\nclassification accuracy of 97.95% and outperforms three state-of-the-art object\nrecognition algorithms, namely - Deep Belief Networks, Convolutional Neural\nNetworks and Stacked Denoising Autoencoders by ~11%. On SAT-6, it produces a\nclassification accuracy of 93.9% and outperforms the other algorithms by ~15%.\nComparative studies with a Random Forest classifier show the advantage of an\nunsupervised learning approach over traditional supervised learning techniques.\nA statistical analysis based on Distribution Separability Criterion and\nIntrinsic Dimensionality Estimation substantiates the effectiveness of our\napproach in learning better representations for satellite imagery.","url_abs":"http://arxiv.org/abs/1509.03602v1","url_pdf":"http://arxiv.org/pdf/1509.03602v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"deepsat-a-learning-framework-for-satellite","repo_url":"https://github.com/debanjanxy/GNR-652","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok"}}],"tasks":[{"task_slug":"classification-1","task_name":"Classification"},{"task_slug":"denoising","task_name":"Denoising"},{"task_slug":"classification","task_name":"General Classification"},{"task_slug":"image-classification","task_name":"Image Classification"},{"task_slug":"object-recognition","task_name":"Object Recognition"},{"task_slug":"satellite-image-classification","task_name":"Satellite Image Classification"},{"task_slug":"image-classification","task_name":"image-classification"}],"methods":[{"method_slug":"deep-belief-network","method_name":"Deep Belief Network"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/satellite-image-classification-on-sat-4","task":"Satellite Image Classification","dataset":"SAT-4","model":"DeepSat","rank_in_archive_order":3,"of":4,"metrics":{"Accuracy":"97.95"},"uses_additional_data":false},{"leaderboard":"/sota/satellite-image-classification-on-sat-4","task":"Satellite Image Classification","dataset":"SAT-4","model":"DBN","rank_in_archive_order":4,"of":4,"metrics":{"Accuracy":"81.78"},"uses_additional_data":false},{"leaderboard":"/sota/satellite-image-classification-on-sat-6","task":"Satellite Image Classification","dataset":"SAT-6","model":"DeepSat","rank_in_archive_order":2,"of":3,"metrics":{"Accuracy":"93.92"},"uses_additional_data":false},{"leaderboard":"/sota/satellite-image-classification-on-sat-6","task":"Satellite Image Classification","dataset":"SAT-6","model":"DBN","rank_in_archive_order":3,"of":3,"metrics":{"Accuracy":"76.47"},"uses_additional_data":false}],"syntology":{"syntology_url":null,"atlas_url":"https://app.syntology.ai/?focus=1509.03602","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}