{"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/endnet-sparse-autoencoder-network-for","title":"EndNet: Sparse AutoEncoder Network for Endmember Extraction and Hyperspectral Unmixing","arxiv_id":"1708.01894","date":"2017-08-06","proceeding":null,"authors":["Savas Ozkan","Berk Kaya","Gozde Bozdagi Akar"],"abstract":"Data acquired from multi-channel sensors is a highly valuable asset to\ninterpret the environment for a variety of remote sensing applications.\nHowever, low spatial resolution is a critical limitation for previous sensors\nand the constituent materials of a scene can be mixed in different fractions\ndue to their spatial interactions. Spectral unmixing is a technique that allows\nus to obtain the material spectral signatures and their fractions from\nhyperspectral data. In this paper, we propose a novel endmember extraction and\nhyperspectral unmixing scheme, so called \\textit{EndNet}, that is based on a\ntwo-staged autoencoder network. This well-known structure is completely\nenhanced and restructured by introducing additional layers and a projection\nmetric (i.e., spectral angle distance (SAD) instead of inner product) to\nachieve an optimum solution. Moreover, we present a novel loss function that is\ncomposed of a Kullback-Leibler divergence term with SAD similarity and\nadditional penalty terms to improve the sparsity of the estimates. These\nmodifications enable us to set the common properties of endmembers such as\nnon-linearity and sparsity for autoencoder networks. Lastly, due to the\nstochastic-gradient based approach, the method is scalable for large-scale data\nand it can be accelerated on Graphical Processing Units (GPUs). To demonstrate\nthe superiority of our proposed method, we conduct extensive experiments on\nseveral well-known datasets. The results confirm that the proposed method\nconsiderably improves the performance compared to the state-of-the-art\ntechniques in literature.","url_abs":"http://arxiv.org/abs/1708.01894v4","url_pdf":"http://arxiv.org/pdf/1708.01894v4.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":"endnet-sparse-autoencoder-network-for","repo_url":"https://github.com/savasozkan/endnet","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"none","reach":{"status":"unanswered"}}],"tasks":[{"task_slug":"hyperspectral-unmixing","task_name":"Hyperspectral Unmixing"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":"https://app.syntology.ai/?focus=1708.01894","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}