{"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/systematic-evaluation-of-cnn-advances-on-the","title":"Systematic evaluation of CNN advances on the ImageNet","arxiv_id":"1606.02228","date":"2016-06-07","proceeding":null,"authors":["Dmytro Mishkin","Nikolay Sergievskiy","Jiri Matas"],"abstract":"The paper systematically studies the impact of a range of recent advances in\nCNN architectures and learning methods on the object categorization (ILSVRC)\nproblem. The evalution tests the influence of the following choices of the\narchitecture: non-linearity (ReLU, ELU, maxout, compatibility with batch\nnormalization), pooling variants (stochastic, max, average, mixed), network\nwidth, classifier design (convolutional, fully-connected, SPP), image\npre-processing, and of learning parameters: learning rate, batch size,\ncleanliness of the data, etc.\n  The performance gains of the proposed modifications are first tested\nindividually and then in combination. The sum of individual gains is bigger\nthan the observed improvement when all modifications are introduced, but the\n\"deficit\" is small suggesting independence of their benefits. We show that the\nuse of 128x128 pixel images is sufficient to make qualitative conclusions about\noptimal network structure that hold for the full size Caffe and VGG nets. The\nresults are obtained an order of magnitude faster than with the standard 224\npixel images.","url_abs":"http://arxiv.org/abs/1606.02228v2","url_pdf":"http://arxiv.org/pdf/1606.02228v2.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":"systematic-evaluation-of-cnn-advances-on-the","repo_url":"https://github.com/ducha-aiki/caffenet-benchmark","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"none","reach":null}],"tasks":[{"task_slug":"object-categorization","task_name":"Object Categorization"}],"methods":[{"method_slug":"convolution","method_name":"Convolution"},{"method_slug":"dense-connections","method_name":"Dense Connections"},{"method_slug":"dropout","method_name":"Dropout"},{"method_slug":"elu","method_name":"ELU"},{"method_slug":"max-pooling","method_name":"Max Pooling"},{"method_slug":"relu","method_name":"ReLU"},{"method_slug":"softmax","method_name":"Softmax"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}