{"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/deepgeo-photo-localization-with-deep-neural","title":"DeepGeo: Photo Localization with Deep Neural Network","arxiv_id":"1810.03077","date":"2018-10-07","proceeding":null,"authors":["Sudharshan Suresh","Nathaniel Chodosh","Montiel Abello"],"abstract":"In this paper we address the task of determining the geographical location of\nan image, a pertinent problem in learning and computer vision. This research\nwas inspired from playing GeoGuessr, a game that tests a humans' ability to\nlocalize themselves using just images of their surroundings. In particular, we\nwish to investigate how geographical, ecological and man-made features\ngeneralize for random location prediction. This is framed as a classification\nproblem: given images sampled from the USA, the most-probable state among 50 is\npredicted. Previous work uses models extensively trained on large, unfiltered\nonline datasets that are primed towards specific locations. To this end, we\ncreate (and open-source) the 50States10K dataset - with 0.5 million Google\nStreet View images of the country. A deep neural network based on the ResNet\narchitecture is trained, and four different strategies of incorporating\nlow-level cardinality information are presented. This model achieves an\naccuracy 20 times better than chance on a test dataset, which rises to 71.87%\nwhen taking the best of top-5 guesses. The network also beats human subjects in\n4 out of 5 rounds of GeoGuessr.","url_abs":"http://arxiv.org/abs/1810.03077v1","url_pdf":"http://arxiv.org/pdf/1810.03077v1.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":"deepgeo-photo-localization-with-deep-neural","repo_url":"https://github.com/suddhu/DeepGeo","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"none","reach":null},{"paper_slug":"deepgeo-photo-localization-with-deep-neural","repo_url":"https://github.com/kvsnoufal/ImageGeoLocation","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/1810.03077","atlas_url":"https://app.syntology.ai/?focus=1810.03077","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}