{"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/clcnet-improving-the-efficiency-of","title":"clcNet: Improving the Efficiency of Convolutional Neural Network using Channel Local Convolutions","arxiv_id":"1712.06145","date":"2017-12-17","proceeding":"CVPR 2018 6","authors":["Dong-Qing Zhang"],"abstract":"Depthwise convolution and grouped convolution has been successfully applied\nto improve the efficiency of convolutional neural network (CNN). We suggest\nthat these models can be considered as special cases of a generalized\nconvolution operation, named channel local convolution(CLC), where an output\nchannel is computed using a subset of the input channels. This definition\nentails computation dependency relations between input and output channels,\nwhich can be represented by a channel dependency graph(CDG). By modifying the\nCDG of grouped convolution, a new CLC kernel named interlaced grouped\nconvolution (IGC) is created. Stacking IGC and GC kernels results in a\nconvolution block (named CLC Block) for approximating regular convolution. By\nresorting to the CDG as an analysis tool, we derive the rule for setting the\nmeta-parameters of IGC and GC and the framework for minimizing the\ncomputational cost. A new CNN model named clcNet is then constructed using CLC\nblocks, which shows significantly higher computational efficiency and fewer\nparameters compared to state-of-the-art networks, when being tested using the\nImageNet-1K dataset. Source code is available at\nhttps://github.com/dqzhang17/clcnet.torch .","url_abs":"http://arxiv.org/abs/1712.06145v3","url_pdf":"http://arxiv.org/pdf/1712.06145v3.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":"clcnet-improving-the-efficiency-of","repo_url":"https://github.com/dqzhang17/clcnet.torch","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"torch","reach":null}],"tasks":[{"task_slug":"computational-efficiency","task_name":"Computational Efficiency"}],"methods":[{"method_slug":"1x1-convolution","method_name":"1x1 Convolution"},{"method_slug":"convolution","method_name":"Convolution"},{"method_slug":"grouped-convolution","method_name":"Grouped Convolution"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}