{"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/adversarial-examples-improve-image","title":"Adversarial Examples Improve Image Recognition","arxiv_id":"1911.09665","date":"2019-11-21","proceeding":"CVPR 2020 6","authors":["Cihang Xie","Mingxing Tan","Boqing Gong","Jiang Wang","Alan Yuille","Quoc V. Le"],"abstract":"Adversarial examples are commonly viewed as a threat to ConvNets. Here we present an opposite perspective: adversarial examples can be used to improve image recognition models if harnessed in the right manner. We propose AdvProp, an enhanced adversarial training scheme which treats adversarial examples as additional examples, to prevent overfitting. Key to our method is the usage of a separate auxiliary batch norm for adversarial examples, as they have different underlying distributions to normal examples. We show that AdvProp improves a wide range of models on various image recognition tasks and performs better when the models are bigger. For instance, by applying AdvProp to the latest EfficientNet-B7 [28] on ImageNet, we achieve significant improvements on ImageNet (+0.7%), ImageNet-C (+6.5%), ImageNet-A (+7.0%), Stylized-ImageNet (+4.8%). With an enhanced EfficientNet-B8, our method achieves the state-of-the-art 85.5% ImageNet top-1 accuracy without extra data. This result even surpasses the best model in [20] which is trained with 3.5B Instagram images (~3000X more than ImageNet) and ~9.4X more parameters. Models are available at https://github.com/tensorflow/tpu/tree/master/models/official/efficientnet.","url_abs":"https://arxiv.org/abs/1911.09665v2","url_pdf":"https://arxiv.org/pdf/1911.09665v2.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":"adversarial-examples-improve-image","repo_url":"https://github.com/tensorflow/tpu","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"tf","reach":{"status":"ok","spdx":"Apache-2.0"}},{"paper_slug":"adversarial-examples-improve-image","repo_url":"https://github.com/rwightman/pytorch-image-models","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"adversarial-examples-improve-image","repo_url":"https://github.com/yaoshiang/MobileNetV2-CIFAR-Cleverhans","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok"}},{"paper_slug":"adversarial-examples-improve-image","repo_url":"https://github.com/open-mmlab/mmclassification","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"pytorch","reach":{"status":"ok","spdx":"Apache-2.0"}},{"paper_slug":"adversarial-examples-improve-image","repo_url":"https://github.com/osmr/imgclsmob","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"mxnet","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"adversarial-examples-improve-image","repo_url":"https://github.com/tingxueronghua/pytorch-classification-advprop","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"domain-generalization","task_name":"Domain Generalization"},{"task_slug":"image-classification","task_name":"Image Classification"}],"methods":[{"method_slug":"1x1-convolution","method_name":"1x1 Convolution"},{"method_slug":"advprop","method_name":"AdvProp"},{"method_slug":"autoaugment","method_name":"AutoAugment"},{"method_slug":"auxiliary-batch-normalization","method_name":"Auxiliary Batch Normalization"},{"method_slug":"average-pooling","method_name":"Average Pooling"},{"method_slug":"batch-normalization","method_name":"Batch Normalization"},{"method_slug":"colorjitter","method_name":"ColorJitter"},{"method_slug":"convolution","method_name":"Convolution"},{"method_slug":"cutout","method_name":"Cutout"},{"method_slug":"dense-connections","method_name":"Dense Connections"},{"method_slug":"depthwise-convolution","method_name":"Depthwise Convolution"},{"method_slug":"depthwise-separable-convolution","method_name":"Depthwise Separable Convolution"},{"method_slug":"dropout","method_name":"Dropout"},{"method_slug":"efficientnet","method_name":"EfficientNet"},{"method_slug":"inverted-residual-block","method_name":"Inverted Residual Block"},{"method_slug":"lstm","method_name":"LSTM"},{"method_slug":"pointwise-convolution","method_name":"Pointwise Convolution"},{"method_slug":"rmsprop","method_name":"RMSProp"},{"method_slug":"relu","method_name":"ReLU"},{"method_slug":"sigmoid-activation","method_name":"Sigmoid Activation"},{"method_slug":"squeeze-and-excitation-block","method_name":"Squeeze-and-Excitation Block"},{"method_slug":"tanh-activation","method_name":"Tanh Activation"},{"method_slug":"weight-decay","method_name":"Weight Decay"}],"datasets_introduced":[],"methods_introduced":[{"slug":"advprop","name":"AdvProp","full_name":"AdvProp"},{"slug":"auxiliary-batch-normalization","name":"Auxiliary Batch Normalization","full_name":"Auxiliary Batch Normalization"}],"results":[{"leaderboard":"/sota/domain-generalization-on-vizwiz","task":"Domain Generalization","dataset":"VizWiz-Classification","model":"EfficientNet-B8 (advprop+autoaug)","rank_in_archive_order":4,"of":90,"metrics":{"Accuracy - All Images":"50.5","Accuracy - Clean Images":"53.2","Accuracy - Corrupted Images":"45.8"},"uses_additional_data":false},{"leaderboard":"/sota/domain-generalization-on-vizwiz","task":"Domain Generalization","dataset":"VizWiz-Classification","model":"EfficientNet-B7 (advprop+autoaug)","rank_in_archive_order":5,"of":90,"metrics":{"Accuracy - All Images":"49.7","Accuracy - Clean Images":"52","Accuracy - Corrupted Images":"45"},"uses_additional_data":false},{"leaderboard":"/sota/domain-generalization-on-vizwiz","task":"Domain Generalization","dataset":"VizWiz-Classification","model":"EfficientNet-B6 (advprop+autoaug)","rank_in_archive_order":6,"of":90,"metrics":{"Accuracy - All Images":"49.6","Accuracy - Clean Images":"53.2","Accuracy - Corrupted Images":"44.7"},"uses_additional_data":false},{"leaderboard":"/sota/domain-generalization-on-vizwiz","task":"Domain Generalization","dataset":"VizWiz-Classification","model":"EfficientNet-B5 (advprop+autoaug)","rank_in_archive_order":7,"of":90,"metrics":{"Accuracy - All Images":"49.1","Accuracy - Clean Images":"51.7","Accuracy - Corrupted Images":"44"},"uses_additional_data":false},{"leaderboard":"/sota/domain-generalization-on-vizwiz","task":"Domain Generalization","dataset":"VizWiz-Classification","model":"EfficientNet-B4 (advprop+autoaug)","rank_in_archive_order":10,"of":90,"metrics":{"Accuracy - All Images":"48.1","Accuracy - Clean Images":"51.4","Accuracy - Corrupted Images":"42.5"},"uses_additional_data":false},{"leaderboard":"/sota/domain-generalization-on-vizwiz","task":"Domain Generalization","dataset":"VizWiz-Classification","model":"EfficientNet-B3 (advprop+autoaug)","rank_in_archive_order":15,"of":90,"metrics":{"Accuracy - All Images":"45.5","Accuracy - Clean Images":"49.5","Accuracy - Corrupted Images":"39.8"},"uses_additional_data":false},{"leaderboard":"/sota/domain-generalization-on-vizwiz","task":"Domain Generalization","dataset":"VizWiz-Classification","model":"EfficientNet-B2 (advprop+autoaug)","rank_in_archive_order":19,"of":90,"metrics":{"Accuracy - All Images":"44.3","Accuracy - Clean Images":"48","Accuracy - Corrupted Images":"38.2"},"uses_additional_data":false},{"leaderboard":"/sota/domain-generalization-on-vizwiz","task":"Domain Generalization","dataset":"VizWiz-Classification","model":"EfficientNet-B1 (advprop+autoaug)","rank_in_archive_order":23,"of":90,"metrics":{"Accuracy - All Images":"42.4","Accuracy - Clean Images":"46.7","Accuracy - Corrupted Images":"36.2"},"uses_additional_data":false},{"leaderboard":"/sota/domain-generalization-on-vizwiz","task":"Domain Generalization","dataset":"VizWiz-Classification","model":"EfficientNet-B0 (advprop+autoaug)","rank_in_archive_order":34,"of":90,"metrics":{"Accuracy - All Images":"40.5","Accuracy - Clean Images":"44.9","Accuracy - Corrupted Images":"34.2"},"uses_additional_data":false},{"leaderboard":"/sota/image-classification-on-imagenet","task":"Image Classification","dataset":"ImageNet","model":"AdvProp (EfficientNet-B8)","rank_in_archive_order":224,"of":1060,"metrics":{"Number of params":"88M","Top 1 Accuracy":"85.5%"},"uses_additional_data":false},{"leaderboard":"/sota/image-classification-on-imagenet","task":"Image Classification","dataset":"ImageNet","model":"AdvProp (EfficientNet-B7)","rank_in_archive_order":249,"of":1060,"metrics":{"Number of params":"66M","Top 1 Accuracy":"85.2%"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1911.09665","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1911.09665"}},"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/rwightman/pytorch-image-models","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/tensorflow/tpu","reach":{"status":"ok","spdx":"Apache-2.0"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/tingxueronghua/pytorch-classification-advprop","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/open-mmlab/mmclassification","reach":{"status":"ok","spdx":"Apache-2.0"}},{"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/yaoshiang/MobileNetV2-CIFAR-Cleverhans","reach":{"status":"ok"}}],"summary":{"ran_draft_wrong":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":0,"samples":[{"code_sha256_prefix":"6a259f7cc6738039","entry":"label_smoothing_cross_entropy","repo":"tingxueronghua/pytorch-classification-advprop","repo_kind":"listed","path":"imagenet.py","file_url":"https://github.com/tingxueronghua/pytorch-classification-advprop/blob/HEAD/imagenet.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"MISDECLARED","metamorphic_tier":"invariant","behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"6a259f7cc6738039"}},{"code_sha256_prefix":"50afa2863ffd9fde","entry":"plot_overlap","repo":"tingxueronghua/pytorch-classification-advprop","repo_kind":"listed","path":"utils/logger.py","file_url":"https://github.com/tingxueronghua/pytorch-classification-advprop/blob/HEAD/utils/logger.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"50afa2863ffd9fde"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}