{"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/delving-deep-into-rectifiers-surpassing-human","title":"Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classification","arxiv_id":"1502.01852","date":"2015-02-06","proceeding":"ICCV 2015 12","authors":["Kaiming He","Xiangyu Zhang","Shaoqing Ren","Jian Sun"],"abstract":"Rectified activation units (rectifiers) are essential for state-of-the-art\nneural networks. In this work, we study rectifier neural networks for image\nclassification from two aspects. First, we propose a Parametric Rectified\nLinear Unit (PReLU) that generalizes the traditional rectified unit. PReLU\nimproves model fitting with nearly zero extra computational cost and little\noverfitting risk. Second, we derive a robust initialization method that\nparticularly considers the rectifier nonlinearities. This method enables us to\ntrain extremely deep rectified models directly from scratch and to investigate\ndeeper or wider network architectures. Based on our PReLU networks\n(PReLU-nets), we achieve 4.94% top-5 test error on the ImageNet 2012\nclassification dataset. This is a 26% relative improvement over the ILSVRC 2014\nwinner (GoogLeNet, 6.66%). To our knowledge, our result is the first to surpass\nhuman-level performance (5.1%, Russakovsky et al.) on this visual recognition\nchallenge.","url_abs":"http://arxiv.org/abs/1502.01852v1","url_pdf":"http://arxiv.org/pdf/1502.01852v1.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":"delving-deep-into-rectifiers-surpassing-human","repo_url":"https://github.com/AnzorGozalishvili/autoencoders_playground","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok"}},{"paper_slug":"delving-deep-into-rectifiers-surpassing-human","repo_url":"https://github.com/LiamLau1/MLDE","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok"}},{"paper_slug":"delving-deep-into-rectifiers-surpassing-human","repo_url":"https://github.com/MrtnMndt/Rethinking_CNN_Layerwise_Feature_Amounts","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"delving-deep-into-rectifiers-surpassing-human","repo_url":"https://github.com/dmbernaal/Daedalus","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}},{"paper_slug":"delving-deep-into-rectifiers-surpassing-human","repo_url":"https://github.com/eatamath/metallic","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}},{"paper_slug":"delving-deep-into-rectifiers-surpassing-human","repo_url":"https://github.com/hamiddimyati/dd2424-deep-learning","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":null},{"paper_slug":"delving-deep-into-rectifiers-surpassing-human","repo_url":"https://github.com/hwalsuklee/tensorflow-mnist-MLP-batch_normalization-weight_initializers","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok"}},{"paper_slug":"delving-deep-into-rectifiers-surpassing-human","repo_url":"https://github.com/ihsuy/Train-by-Reconnect","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":null},{"paper_slug":"delving-deep-into-rectifiers-surpassing-human","repo_url":"https://github.com/krish-pinninti/api-ann-python","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok"}},{"paper_slug":"delving-deep-into-rectifiers-surpassing-human","repo_url":"https://github.com/patconrey/ANN-Example","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok"}},{"paper_slug":"delving-deep-into-rectifiers-surpassing-human","repo_url":"https://github.com/phogbinh/handwritten-digit-recognition","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"delving-deep-into-rectifiers-surpassing-human","repo_url":"https://github.com/salluru007/papers","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok"}},{"paper_slug":"delving-deep-into-rectifiers-surpassing-human","repo_url":"https://github.com/sudhirk999/ResearchPapersLinks","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok"}},{"paper_slug":"delving-deep-into-rectifiers-surpassing-human","repo_url":"https://github.com/tanjeffreyz/deep-residual-learning","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"delving-deep-into-rectifiers-surpassing-human","repo_url":"https://github.com/zonetrooper32/VDCNN","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok"}}],"tasks":[{"task_slug":"classification","task_name":"General Classification"},{"task_slug":"image-classification","task_name":"Image Classification"},{"task_slug":"image-classification","task_name":"image-classification"}],"methods":[{"method_slug":"colorjitter","method_name":"ColorJitter"},{"method_slug":"convolution","method_name":"Convolution"},{"method_slug":"dense-connections","method_name":"Dense Connections"},{"method_slug":"dropout","method_name":"Dropout"},{"method_slug":"kaiming-initialization","method_name":"Kaiming Initialization"},{"method_slug":"max-pooling","method_name":"Max Pooling"},{"method_slug":"prelu","method_name":"PReLU"},{"method_slug":"prelu-net","method_name":"PReLU-Net"},{"method_slug":"randomhorizontalflip","method_name":"Random Horizontal Flip"},{"method_slug":"random-resized-crop","method_name":"Random Resized Crop"},{"method_slug":"sgd-with-momentum","method_name":"SGD with Momentum"},{"method_slug":"softmax","method_name":"Softmax"},{"method_slug":"spatial-pyramid-pooling","method_name":"Spatial Pyramid Pooling"},{"method_slug":"step-decay","method_name":"Step Decay"},{"method_slug":"weight-decay","method_name":"Weight Decay"}],"datasets_introduced":[],"methods_introduced":[{"slug":"prelu","name":"PReLU","full_name":"Parameterized ReLU"}],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1502.01852","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1502.01852"}},"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/ihsuy/Train-by-Reconnect","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/patconrey/ANN-Example","reach":{"status":"ok"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/zonetrooper32/VDCNN","reach":{"status":"ok"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/salluru007/papers","reach":{"status":"ok"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/dmbernaal/Daedalus","reach":{"status":"unanswered"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/hamiddimyati/dd2424-deep-learning","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/MrtnMndt/Rethinking_CNN_Layerwise_Feature_Amounts","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/AnzorGozalishvili/autoencoders_playground","reach":{"status":"ok"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/hwalsuklee/tensorflow-mnist-MLP-batch_normalization-weight_initializers","reach":{"status":"ok"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/phogbinh/handwritten-digit-recognition","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/krish-pinninti/api-ann-python","reach":{"status":"ok"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/eatamath/metallic","reach":{"status":"ok"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/sudhirk999/ResearchPapersLinks","reach":{"status":"ok"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/LiamLau1/MLDE","reach":{"status":"ok"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/tanjeffreyz/deep-residual-learning","reach":null}],"summary":{"ran_honours":1},"by_repo_kind":{"listed":{"samples":1,"ran":1,"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":1,"samples":[{"code_sha256_prefix":"94d5729d92a97933","entry":"relu","repo":"hamiddimyati/dd2424-deep-learning","repo_kind":"listed","path":"assignment-2/assignment2-bonus.py","file_url":"https://github.com/hamiddimyati/dd2424-deep-learning/blob/HEAD/assignment-2/assignment2-bonus.py","link_basis":"first_harvest_node","language":"python","status":"ran_honours","verification_level":1,"contract_check":"HONOURS","metamorphic_tier":"invariant","behaviour_fingerprint":true,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"94d5729d92a97933"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}