{"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/pedestrian-detection-inspired-by-appearance","title":"Pedestrian Detection Inspired by Appearance Constancy and Shape Symmetry","arxiv_id":"1511.08058","date":"2015-11-25","proceeding":"CVPR 2016 6","authors":["Jiale Cao","Yanwei Pang","Xuelong. Li"],"abstract":"The discrimination and simplicity of features are very important for\neffective and efficient pedestrian detection. However, most state-of-the-art\nmethods are unable to achieve good tradeoff between accuracy and efficiency.\nInspired by some simple inherent attributes of pedestrians (i.e., appearance\nconstancy and shape symmetry), we propose two new types of non-neighboring\nfeatures (NNF): side-inner difference features (SIDF) and symmetrical\nsimilarity features (SSF). SIDF can characterize the difference between the\nbackground and pedestrian and the difference between the pedestrian contour and\nits inner part. SSF can capture the symmetrical similarity of pedestrian shape.\nHowever, it's difficult for neighboring features to have such above\ncharacterization abilities. Finally, we propose to combine both non-neighboring\nand neighboring features for pedestrian detection. It's found that\nnon-neighboring features can further decrease the average miss rate by 4.44%.\nExperimental results on INRIA and Caltech pedestrian datasets demonstrate the\neffectiveness and efficiency of the proposed method. Compared to the\nstate-of-the-art methods without using CNN, our method achieves the best\ndetection performance on Caltech, outperforming the second best method (i.e.,\nCheckboards) by 1.63%.","url_abs":"http://arxiv.org/abs/1511.08058v1","url_pdf":"http://arxiv.org/pdf/1511.08058v1.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":"pedestrian-detection","task_name":"Pedestrian Detection"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/pedestrian-detection-on-caltech","task":"Pedestrian Detection","dataset":"Caltech","model":"NNNF","rank_in_archive_order":28,"of":33,"metrics":{"Reasonable Miss Rate":"16.20"},"uses_additional_data":false}],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}