{"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/optical-flow-dataset-and-benchmark-for-visual","title":"Optical Flow Dataset and Benchmark for Visual Crowd Analysis","arxiv_id":"1811.07170","date":"2018-11-17","proceeding":null,"authors":["Gregory Schröder","Tobias Senst","Erik Bochinski","Thomas Sikora"],"abstract":"The performance of optical flow algorithms greatly depends on the specifics\nof the content and the application for which it is used. Existing and well\nestablished optical flow datasets are limited to rather particular contents\nfrom which none is close to crowd behavior analysis; whereas such applications\nheavily utilize optical flow. We introduce a new optical flow dataset\nexploiting the possibilities of a recent video engine to generate sequences\nwith ground-truth optical flow for large crowds in different scenarios. We\nbreak with the development of the last decade of introducing ever increasing\ndisplacements to pose new difficulties. Instead we focus on real-world\nsurveillance scenarios where numerous small, partly independent, non rigidly\nmoving objects observed over a long temporal range pose a challenge. By\nevaluating different optical flow algorithms, we find that results of\nestablished datasets can not be transferred to these new challenges. In\nexhaustive experiments we are able to provide new insight into optical flow for\ncrowd analysis. Finally, the results have been validated on the real-world UCF\ncrowd tracking benchmark while achieving competitive results compared to more\nsophisticated state-of-the-art crowd tracking approaches.","url_abs":"http://arxiv.org/abs/1811.07170v1","url_pdf":"http://arxiv.org/pdf/1811.07170v1.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":"optical-flow-dataset-and-benchmark-for-visual","repo_url":"https://github.com/tsenst/CrowdFlow","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"none","reach":null}],"tasks":[{"task_slug":"optical-flow-estimation","task_name":"Optical Flow Estimation"},{"task_slug":"visual-crowd-analysis","task_name":"Visual Crowd Analysis"}],"methods":[],"datasets_introduced":[{"slug":"crowdflow","name":"CrowdFlow","full_name":"TUB CrowdFlow"}],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}