{"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/a-generative-map-for-image-based-camera","title":"A Generative Map for Image-based Camera Localization","arxiv_id":"1902.11124","date":"2019-02-18","proceeding":null,"authors":["Mingpan Guo","Stefan Matthes","Jiaojiao Ye","Hao Shen"],"abstract":"In image-based camera localization systems, information about the environment\nis usually stored in some representation, which can be referred to as a map.\nConventionally, most maps are built upon hand-crafted features. Recently,\nneural networks have attracted attention as a data-driven map representation,\nand have shown promising results in visual localization. However, these neural\nnetwork maps are generally hard to interpret by human. A readable map is not\nonly accessible to humans, but also provides a way to be verified when the\nground truth pose is unavailable. To tackle this problem, we propose Generative\nMap, a new framework for learning human-readable neural network maps, by\ncombining a generative model with the Kalman filter, which also allows it to\nincorporate additional sensor information such as stereo visual odometry. For\nevaluation, we use real world images from the 7-Scenes and Oxford RobotCar\ndatasets. We demonstrate that our Generative Map can be queried with a pose of\ninterest from the test sequence to predict an image, which closely resembles\nthe true scene. For localization, we show that Generative Map achieves\ncomparable performance with current regression models. Moreover, our framework\nis trained completely from scratch, unlike regression models which rely on\nlarge ImageNet pretrained networks.","url_abs":"http://arxiv.org/abs/1902.11124v4","url_pdf":"http://arxiv.org/pdf/1902.11124v4.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":"a-generative-map-for-image-based-camera","repo_url":"https://github.com/Mingpan/generative_map","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"tf","reach":null}],"tasks":[{"task_slug":"camera-localization","task_name":"Camera Localization"},{"task_slug":"visual-localization","task_name":"Visual Localization"},{"task_slug":"visual-odometry","task_name":"Visual Odometry"},{"task_slug":"regression-1","task_name":"regression"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}