{"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-detection-of-people-and-their-mobility","title":"Deep Detection of People and their Mobility Aids for a Hospital Robot","arxiv_id":"1708.00674","date":"2017-08-02","proceeding":null,"authors":["Andres Vasquez","Marina Kollmitz","Andreas Eitel","Wolfram Burgard"],"abstract":"Robots operating in populated environments encounter many different types of\npeople, some of whom might have an advanced need for cautious interaction,\nbecause of physical impairments or their advanced age. Robots therefore need to\nrecognize such advanced demands to provide appropriate assistance, guidance or\nother forms of support. In this paper, we propose a depth-based perception\npipeline that estimates the position and velocity of people in the environment\nand categorizes them according to the mobility aids they use: pedestrian,\nperson in wheelchair, person in a wheelchair with a person pushing them, person\nwith crutches and person using a walker. We present a fast region proposal\nmethod that feeds a Region-based Convolutional Network (Fast R-CNN). With this,\nwe speed up the object detection process by a factor of seven compared to a\ndense sliding window approach. We furthermore propose a probabilistic position,\nvelocity and class estimator to smooth the CNN's detections and account for\nocclusions and misclassifications. In addition, we introduce a new hospital\ndataset with over 17,000 annotated RGB-D images. Extensive experiments confirm\nthat our pipeline successfully keeps track of people and their mobility aids,\neven in challenging situations with multiple people from different categories\nand frequent occlusions. Videos of our experiments and the dataset are\navailable at http://www2.informatik.uni-freiburg.de/~kollmitz/MobilityAids","url_abs":"http://arxiv.org/abs/1708.00674v1","url_pdf":"http://arxiv.org/pdf/1708.00674v1.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":[],"tasks":[{"task_slug":"object-detection","task_name":"Object Detection"},{"task_slug":null,"task_name":"Position"},{"task_slug":"region-proposal","task_name":"Region Proposal"},{"task_slug":"object-detection-1","task_name":"object-detection"}],"methods":[{"method_slug":"speed","method_name":"SPEED"}],"datasets_introduced":[{"slug":"mobilityaids","name":"MobilityAids","full_name":"MobilityAids"}],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}