{"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/constructing-fast-network-through","title":"Constructing Fast Network through Deconstruction of Convolution","arxiv_id":"1806.07370","date":"2018-05-28","proceeding":"NeurIPS 2018 12","authors":["Yunho Jeon","Junmo Kim"],"abstract":"Convolutional neural networks have achieved great success in various vision\ntasks; however, they incur heavy resource costs. By using deeper and wider\nnetworks, network accuracy can be improved rapidly. However, in an environment\nwith limited resources (e.g., mobile applications), heavy networks may not be\nusable. This study shows that naive convolution can be deconstructed into a\nshift operation and pointwise convolution. To cope with various convolutions,\nwe propose a new shift operation called active shift layer (ASL) that\nformulates the amount of shift as a learnable function with shift parameters.\nThis new layer can be optimized end-to-end through backpropagation and it can\nprovide optimal shift values. Finally, we apply this layer to a light and fast\nnetwork that surpasses existing state-of-the-art networks.","url_abs":"http://arxiv.org/abs/1806.07370v5","url_pdf":"http://arxiv.org/pdf/1806.07370v5.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":"constructing-fast-network-through","repo_url":"https://github.com/jyh2986/Active-Shift","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok","spdx":"NOASSERTION"}},{"paper_slug":"constructing-fast-network-through","repo_url":"https://github.com/DeadAt0m/ActiveSparseShifts-PyTorch","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}}],"tasks":[],"methods":[{"method_slug":"convolution","method_name":"Convolution"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1806.07370","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}