{"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/pruning-convolutional-neural-networks-for","title":"Pruning Convolutional Neural Networks for Resource Efficient Inference","arxiv_id":"1611.06440","date":"2016-11-19","proceeding":null,"authors":["Pavlo Molchanov","Stephen Tyree","Tero Karras","Timo Aila","Jan Kautz"],"abstract":"We propose a new formulation for pruning convolutional kernels in neural\nnetworks to enable efficient inference. We interleave greedy criteria-based\npruning with fine-tuning by backpropagation - a computationally efficient\nprocedure that maintains good generalization in the pruned network. We propose\na new criterion based on Taylor expansion that approximates the change in the\ncost function induced by pruning network parameters. We focus on transfer\nlearning, where large pretrained networks are adapted to specialized tasks. The\nproposed criterion demonstrates superior performance compared to other\ncriteria, e.g. the norm of kernel weights or feature map activation, for\npruning large CNNs after adaptation to fine-grained classification tasks\n(Birds-200 and Flowers-102) relaying only on the first order gradient\ninformation. We also show that pruning can lead to more than 10x theoretical\n(5x practical) reduction in adapted 3D-convolutional filters with a small drop\nin accuracy in a recurrent gesture classifier. Finally, we show results for the\nlarge-scale ImageNet dataset to emphasize the flexibility of our approach.","url_abs":"http://arxiv.org/abs/1611.06440v2","url_pdf":"http://arxiv.org/pdf/1611.06440v2.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":"pruning-convolutional-neural-networks-for","repo_url":"https://github.com/EkdeepSLubana/OrthoReg","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"pruning-convolutional-neural-networks-for","repo_url":"https://github.com/azhe198827/channel_prune","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"pruning-convolutional-neural-networks-for","repo_url":"https://github.com/dongkwan-kim/Adaptive-Forgetting","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok"}},{"paper_slug":"pruning-convolutional-neural-networks-for","repo_url":"https://github.com/eeric/channel_prune","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}},{"paper_slug":"pruning-convolutional-neural-networks-for","repo_url":"https://github.com/kcang2/Prune_U-Net","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"GPL-3.0"}},{"paper_slug":"pruning-convolutional-neural-networks-for","repo_url":"https://github.com/matthew-mcateer/Keras_pruning","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok"}},{"paper_slug":"pruning-convolutional-neural-networks-for","repo_url":"https://github.com/rohinarora/Deep-Learning-Papers","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok"}},{"paper_slug":"pruning-convolutional-neural-networks-for","repo_url":"https://github.com/zepx/pytorch-weight-prune","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}}],"tasks":[{"task_slug":"transfer-learning","task_name":"Transfer Learning"}],"methods":[{"method_slug":"pruning","method_name":"Pruning"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1611.06440","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1611.06440"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/dongkwan-kim/Adaptive-Forgetting","reach":{"status":"ok"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/zepx/pytorch-weight-prune","reach":{"status":"ok"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/kcang2/Prune_U-Net","reach":{"status":"ok","spdx":"GPL-3.0"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/EkdeepSLubana/OrthoReg","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/matthew-mcateer/Keras_pruning","reach":{"status":"ok"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/rohinarora/Deep-Learning-Papers","reach":{"status":"ok"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/eeric/channel_prune","reach":{"status":"ok"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/azhe198827/channel_prune","reach":null}],"summary":{"ran_fixture":2},"by_repo_kind":{"listed":{"samples":2,"ran":2,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":2,"samples":[{"code_sha256_prefix":"543b868f099e65b8","entry":"prune_resnet18_conv_layer","repo":"azhe198827/channel_prune","repo_kind":"listed","path":"prune.py","file_url":"https://github.com/azhe198827/channel_prune/blob/HEAD/prune.py","link_basis":"first_harvest_node","language":"python","status":"ran_fixture","verification_level":1,"contract_check":"RAISES","metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"543b868f099e65b8"}},{"code_sha256_prefix":"42fb20662693ce7e","entry":"replace_layers","repo":"azhe198827/channel_prune","repo_kind":"listed","path":"prune.py","file_url":"https://github.com/azhe198827/channel_prune/blob/HEAD/prune.py","link_basis":"first_harvest_node","language":"python","status":"ran_fixture","verification_level":1,"contract_check":"RAISES","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"42fb20662693ce7e"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}