{"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/generalized-orderless-pooling-performs","title":"Generalized orderless pooling performs implicit salient matching","arxiv_id":"1705.00487","date":"2017-05-01","proceeding":"ICCV 2017 10","authors":["Marcel Simon","Yang Gao","Trevor Darrell","Joachim Denzler","Erik Rodner"],"abstract":"Most recent CNN architectures use average pooling as a final feature encoding\nstep. In the field of fine-grained recognition, however, recent global\nrepresentations like bilinear pooling offer improved performance. In this\npaper, we generalize average and bilinear pooling to \"alpha-pooling\", allowing\nfor learning the pooling strategy during training. In addition, we present a\nnovel way to visualize decisions made by these approaches. We identify parts of\ntraining images having the highest influence on the prediction of a given test\nimage. It allows for justifying decisions to users and also for analyzing the\ninfluence of semantic parts. For example, we can show that the higher capacity\nVGG16 model focuses much more on the bird's head than, e.g., the lower-capacity\nVGG-M model when recognizing fine-grained bird categories. Both contributions\nallow us to analyze the difference when moving between average and bilinear\npooling. In addition, experiments show that our generalized approach can\noutperform both across a variety of standard datasets.","url_abs":"http://arxiv.org/abs/1705.00487v3","url_pdf":"http://arxiv.org/pdf/1705.00487v3.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":"generalized-orderless-pooling-performs","repo_url":"https://github.com/cvjena/alpha_pooling","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"caffe2","reach":null},{"paper_slug":"generalized-orderless-pooling-performs","repo_url":"https://github.com/KeremTurgutlu/bcnn","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":null}],"tasks":[],"methods":[{"method_slug":"average-pooling","method_name":"Average Pooling"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}