{"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/sharesnet-reducing-residual-network-parameter","title":"ShaResNet: reducing residual network parameter number by sharing weights","arxiv_id":"1702.08782","date":"2017-02-28","proceeding":null,"authors":["Alexandre Boulch"],"abstract":"Deep Residual Networks have reached the state of the art in many image\nprocessing tasks such image classification. However, the cost for a gain in\naccuracy in terms of depth and memory is prohibitive as it requires a higher\nnumber of residual blocks, up to double the initial value. To tackle this\nproblem, we propose in this paper a way to reduce the redundant information of\nthe networks. We share the weights of convolutional layers between residual\nblocks operating at the same spatial scale. The signal flows multiple times in\nthe same convolutional layer. The resulting architecture, called ShaResNet,\ncontains block specific layers and shared layers. These ShaResNet are trained\nexactly in the same fashion as the commonly used residual networks. We show, on\nthe one hand, that they are almost as efficient as their sequential\ncounterparts while involving less parameters, and on the other hand that they\nare more efficient than a residual network with the same number of parameters.\nFor example, a 152-layer-deep residual network can be reduced to 106\nconvolutional layers, i.e. a parameter gain of 39\\%, while loosing less than\n0.2\\% accuracy on ImageNet.","url_abs":"http://arxiv.org/abs/1702.08782v2","url_pdf":"http://arxiv.org/pdf/1702.08782v2.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":"sharesnet-reducing-residual-network-parameter","repo_url":"https://github.com/aboulch/sharesnet","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"torch","reach":{"status":"ok","spdx":"GPL-3.0"}},{"paper_slug":"sharesnet-reducing-residual-network-parameter","repo_url":"https://github.com/osmr/imgclsmob","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"mxnet","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"image-classification","task_name":"Image Classification"},{"task_slug":"image-classification","task_name":"image-classification"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}