{"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/centripetal-sgd-for-pruning-very-deep","title":"Centripetal SGD for Pruning Very Deep Convolutional Networks with Complicated Structure","arxiv_id":"1904.03837","date":"2019-04-08","proceeding":"CVPR 2019 6","authors":["Xiaohan Ding","Guiguang Ding","Yuchen Guo","Jungong Han"],"abstract":"The redundancy is widely recognized in Convolutional Neural Networks (CNNs),\nwhich enables to remove unimportant filters from convolutional layers so as to\nslim the network with acceptable performance drop. Inspired by the linear and\ncombinational properties of convolution, we seek to make some filters\nincreasingly close and eventually identical for network slimming. To this end,\nwe propose Centripetal SGD (C-SGD), a novel optimization method, which can\ntrain several filters to collapse into a single point in the parameter\nhyperspace. When the training is completed, the removal of the identical\nfilters can trim the network with NO performance loss, thus no finetuning is\nneeded. By doing so, we have partly solved an open problem of constrained\nfilter pruning on CNNs with complicated structure, where some layers must be\npruned following others. Our experimental results on CIFAR-10 and ImageNet have\njustified the effectiveness of C-SGD-based filter pruning. Moreover, we have\nprovided empirical evidences for the assumption that the redundancy in deep\nneural networks helps the convergence of training by showing that a redundant\nCNN trained using C-SGD outperforms a normally trained counterpart with the\nequivalent width.","url_abs":"http://arxiv.org/abs/1904.03837v1","url_pdf":"http://arxiv.org/pdf/1904.03837v1.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":"centripetal-sgd-for-pruning-very-deep","repo_url":"https://github.com/ShawnDing1994/Centripetal-SGD","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok","spdx":"Apache-2.0"}}],"tasks":[],"methods":[{"method_slug":"pruning","method_name":"Pruning"},{"method_slug":"sgd","method_name":"SGD"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":"https://app.syntology.ai/?focus=1904.03837","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}