{"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/posenet-a-convolutional-network-for-real-time","title":"PoseNet: A Convolutional Network for Real-Time 6-DOF Camera Relocalization","arxiv_id":"1505.07427","date":"2015-05-27","proceeding":"ICCV 2015 12","authors":["Alex Kendall","Matthew Grimes","Roberto Cipolla"],"abstract":"We present a robust and real-time monocular six degree of freedom\nrelocalization system. Our system trains a convolutional neural network to\nregress the 6-DOF camera pose from a single RGB image in an end-to-end manner\nwith no need of additional engineering or graph optimisation. The algorithm can\noperate indoors and outdoors in real time, taking 5ms per frame to compute. It\nobtains approximately 2m and 6 degree accuracy for large scale outdoor scenes\nand 0.5m and 10 degree accuracy indoors. This is achieved using an efficient 23\nlayer deep convnet, demonstrating that convnets can be used to solve\ncomplicated out of image plane regression problems. This was made possible by\nleveraging transfer learning from large scale classification data. We show the\nconvnet localizes from high level features and is robust to difficult lighting,\nmotion blur and different camera intrinsics where point based SIFT registration\nfails. Furthermore we show how the pose feature that is produced generalizes to\nother scenes allowing us to regress pose with only a few dozen training\nexamples. PoseNet code, dataset and an online demonstration is available on our\nproject webpage, at http://mi.eng.cam.ac.uk/projects/relocalisation/","url_abs":"http://arxiv.org/abs/1505.07427v4","url_pdf":"http://arxiv.org/pdf/1505.07427v4.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":"posenet-a-convolutional-network-for-real-time","repo_url":"https://github.com/alexgkendall/caffe-posenet","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok","spdx":"NOASSERTION"}},{"paper_slug":"posenet-a-convolutional-network-for-real-time","repo_url":"https://github.com/alililia/ascend_posenet","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"mindspore","reach":null},{"paper_slug":"posenet-a-convolutional-network-for-real-time","repo_url":"https://github.com/alililia/gpu_posenet","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"mindspore","reach":null},{"paper_slug":"posenet-a-convolutional-network-for-real-time","repo_url":"https://github.com/code-implementation1/Code6/tree/main/PoseNet","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"mindspore","reach":null},{"paper_slug":"posenet-a-convolutional-network-for-real-time","repo_url":"https://github.com/ertsfftt/posenet","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"mindspore","reach":null},{"paper_slug":"posenet-a-convolutional-network-for-real-time","repo_url":"https://github.com/mindspore-ai/models/blob/master/official/cv/posenet/","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"mindspore","reach":null}],"tasks":[{"task_slug":"camera-relocalization","task_name":"Camera Relocalization"},{"task_slug":"transfer-learning","task_name":"Transfer Learning"}],"methods":[],"datasets_introduced":[{"slug":"cambridge-landmarks","name":"Cambridge Landmarks","full_name":""}],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}