{"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/multi-scale-dense-networks-for-resource","title":"Multi-Scale Dense Networks for Resource Efficient Image Classification","arxiv_id":"1703.09844","date":"2017-03-29","proceeding":"ICLR 2018 1","authors":["Gao Huang","Danlu Chen","Tianhong Li","Felix Wu","Laurens van der Maaten","Kilian Q. Weinberger"],"abstract":"In this paper we investigate image classification with computational resource\nlimits at test time. Two such settings are: 1. anytime classification, where\nthe network's prediction for a test example is progressively updated,\nfacilitating the output of a prediction at any time; and 2. budgeted batch\nclassification, where a fixed amount of computation is available to classify a\nset of examples that can be spent unevenly across \"easier\" and \"harder\" inputs.\nIn contrast to most prior work, such as the popular Viola and Jones algorithm,\nour approach is based on convolutional neural networks. We train multiple\nclassifiers with varying resource demands, which we adaptively apply during\ntest time. To maximally re-use computation between the classifiers, we\nincorporate them as early-exits into a single deep convolutional neural network\nand inter-connect them with dense connectivity. To facilitate high quality\nclassification early on, we use a two-dimensional multi-scale network\narchitecture that maintains coarse and fine level features all-throughout the\nnetwork. Experiments on three image-classification tasks demonstrate that our\nframework substantially improves the existing state-of-the-art in both\nsettings.","url_abs":"http://arxiv.org/abs/1703.09844v5","url_pdf":"http://arxiv.org/pdf/1703.09844v5.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":"multi-scale-dense-networks-for-resource","repo_url":"https://github.com/gaohuang/MSDNet","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"multi-scale-dense-networks-for-resource","repo_url":"https://github.com/gorakraj/earlyexit_onnx","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"multi-scale-dense-networks-for-resource","repo_url":"https://github.com/kalviny/MSDNet-PyTorch","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"multi-scale-dense-networks-for-resource","repo_url":"https://github.com/leaplabthu/l2w-den","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}},{"paper_slug":"multi-scale-dense-networks-for-resource","repo_url":"https://github.com/metodj/anytimeclassification","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"AGPL-3.0"}},{"paper_slug":"multi-scale-dense-networks-for-resource","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"}},{"paper_slug":"multi-scale-dense-networks-for-resource","repo_url":"https://github.com/wszenic/MSD_pytorch","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}}],"tasks":[{"task_slug":"classification","task_name":"General Classification"},{"task_slug":"handwritten-mathmatical-expression","task_name":"Handwritten Mathmatical Expression Recognition"},{"task_slug":"image-classification","task_name":"Image Classification"},{"task_slug":"image-classification","task_name":"image-classification"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/handwritten-mathmatical-expression-1","task":"Handwritten Mathmatical Expression Recognition","dataset":"CROHME 2016","model":"DenseWAP-MSA","rank_in_archive_order":13,"of":16,"metrics":{"ExpRate":"50.1"},"uses_additional_data":false},{"leaderboard":"/sota/handwritten-mathmatical-expression-1","task":"Handwritten Mathmatical Expression Recognition","dataset":"CROHME 2016","model":"DenseWAP","rank_in_archive_order":15,"of":16,"metrics":{"ExpRate":"47.5"},"uses_additional_data":false},{"leaderboard":"/sota/handwritten-mathmatical-expression-2","task":"Handwritten Mathmatical Expression Recognition","dataset":"CROHME 2019","model":"DenseWAP-MSA","rank_in_archive_order":14,"of":14,"metrics":{"ExpRate":"47.7"},"uses_additional_data":false},{"leaderboard":"/sota/handwritten-mathmatical-expression-3","task":"Handwritten Mathmatical Expression Recognition","dataset":"HME100K","model":"DenseWAP","rank_in_archive_order":12,"of":12,"metrics":{"ExpRate":"61.85"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1703.09844","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1703.09844"}},"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/kalviny/MSDNet-PyTorch","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/metodj/anytimeclassification","reach":{"status":"ok","spdx":"AGPL-3.0"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/gaohuang/MSDNet","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/osmr/imgclsmob","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/gorakraj/earlyexit_onnx","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/leaplabthu/l2w-den","reach":{"status":"ok"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/wszenic/MSD_pytorch","reach":{"status":"ok"}}],"summary":{"unverified":4},"by_repo_kind":{"listed":{"samples":4,"ran":0,"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":0,"samples":[{"code_sha256_prefix":"b51a9a4dd721bcbc","entry":"get_dataloaders","repo":"kalviny/MSDNet-PyTorch","repo_kind":"listed","path":"dataloader.py","file_url":"https://github.com/kalviny/MSDNet-PyTorch/blob/HEAD/dataloader.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"b51a9a4dd721bcbc"}},{"code_sha256_prefix":"35a5cf7271901d9c","entry":"get_layer_info","repo":"kalviny/MSDNet-PyTorch","repo_kind":"listed","path":"op_counter.py","file_url":"https://github.com/kalviny/MSDNet-PyTorch/blob/HEAD/op_counter.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"35a5cf7271901d9c"}},{"code_sha256_prefix":"76cc74efee42ba21","entry":"get_num_gen","repo":"kalviny/MSDNet-PyTorch","repo_kind":"listed","path":"op_counter.py","file_url":"https://github.com/kalviny/MSDNet-PyTorch/blob/HEAD/op_counter.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"76cc74efee42ba21"}},{"code_sha256_prefix":"26ff085b343fa39e","entry":"is_leaf","repo":"kalviny/MSDNet-PyTorch","repo_kind":"listed","path":"op_counter.py","file_url":"https://github.com/kalviny/MSDNet-PyTorch/blob/HEAD/op_counter.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"26ff085b343fa39e"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}