{"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/deep-person-detection-in-2d-range-data","title":"Deep Person Detection in 2D Range Data","arxiv_id":"1804.02463","date":"2018-04-06","proceeding":null,"authors":["Lucas Beyer","Alexander Hermans","Timm Linder","Kai O. Arras","Bastian Leibe"],"abstract":"Detecting humans is a key skill for mobile robots and intelligent vehicles in\na large variety of applications. While the problem is well studied for certain\nsensory modalities such as image data, few works exist that address this\ndetection task using 2D range data. However, a widespread sensory setup for\nmany mobile robots in service and domestic applications contains a horizontally\nmounted 2D laser scanner. Detecting people from 2D range data is challenging\ndue to the speed and dynamics of human leg motion and the high levels of\nocclusion and self-occlusion particularly in crowds of people. While previous\napproaches mostly relied on handcrafted features, we recently developed the\ndeep learning based wheelchair and walker detector DROW. In this paper, we show\nthe generalization to people, including small modifications that significantly\nboost DROW's performance. Additionally, by providing a small, fully online\ntemporal window in our network, we further boost our score. We extend the DROW\ndataset with person annotations, making this the largest dataset of person\nannotations in 2D range data, recorded during several days in a real-world\nenvironment with high diversity. Extensive experiments with three current\nbaseline methods indicate it is a challenging dataset, on which our improved\nDROW detector beats the current state-of-the-art.","url_abs":"http://arxiv.org/abs/1804.02463v1","url_pdf":"http://arxiv.org/pdf/1804.02463v1.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":"deep-person-detection-in-2d-range-data","repo_url":"https://github.com/VisualComputingInstitute/2D_lidar_person_detection","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"human-detection","task_name":"Human Detection"}],"methods":[{"method_slug":"speed","method_name":"SPEED"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}