{"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/random-erasing-data-augmentation","title":"Random Erasing Data Augmentation","arxiv_id":"1708.04896","date":"2017-08-16","proceeding":null,"authors":["Zhun Zhong","Liang Zheng","Guoliang Kang","Shaozi Li","Yi Yang"],"abstract":"In this paper, we introduce Random Erasing, a new data augmentation method\nfor training the convolutional neural network (CNN). In training, Random\nErasing randomly selects a rectangle region in an image and erases its pixels\nwith random values. In this process, training images with various levels of\nocclusion are generated, which reduces the risk of over-fitting and makes the\nmodel robust to occlusion. Random Erasing is parameter learning free, easy to\nimplement, and can be integrated with most of the CNN-based recognition models.\nAlbeit simple, Random Erasing is complementary to commonly used data\naugmentation techniques such as random cropping and flipping, and yields\nconsistent improvement over strong baselines in image classification, object\ndetection and person re-identification. Code is available at:\nhttps://github.com/zhunzhong07/Random-Erasing.","url_abs":"http://arxiv.org/abs/1708.04896v2","url_pdf":"http://arxiv.org/pdf/1708.04896v2.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":"random-erasing-data-augmentation","repo_url":"https://github.com/zhunzhong07/Random-Erasing","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}},{"paper_slug":"random-erasing-data-augmentation","repo_url":"https://github.com/Brunogomes97/Imdb","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok"}},{"paper_slug":"random-erasing-data-augmentation","repo_url":"https://github.com/CoinCheung/SphereReID","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"random-erasing-data-augmentation","repo_url":"https://github.com/IyatomiLab/CE-CLCNN","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}},{"paper_slug":"random-erasing-data-augmentation","repo_url":"https://github.com/Janghyeonwoong/Random-Erasing-Tensorflow2","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"unanswered"}},{"paper_slug":"random-erasing-data-augmentation","repo_url":"https://github.com/KentaItakura/CNN-classification-using-random-erasing-and-cut-out","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"unanswered"}},{"paper_slug":"random-erasing-data-augmentation","repo_url":"https://github.com/aditya30394/Person-Re-Identification","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}},{"paper_slug":"random-erasing-data-augmentation","repo_url":"https://github.com/rickyHong/Random-erasing-Data-Augumentation-repl","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}},{"paper_slug":"random-erasing-data-augmentation","repo_url":"https://github.com/rlagywns0213/cifar100_data_augmentation","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}},{"paper_slug":"random-erasing-data-augmentation","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":"random-erasing-data-augmentation","repo_url":"https://github.com/sek788432/Microsoft-Cats-and-Dogs-Image-Classification","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"unanswered"}},{"paper_slug":"random-erasing-data-augmentation","repo_url":"https://github.com/shawnyuen/DataAugmentationPaperCollection","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok"}},{"paper_slug":"random-erasing-data-augmentation","repo_url":"https://github.com/NVlabs/DG-Net","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"pytorch","reach":{"status":"ok","spdx":"NOASSERTION"}},{"paper_slug":"random-erasing-data-augmentation","repo_url":"https://github.com/PaddlePaddle/PaddleClas","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"paddle","reach":{"status":"ok","spdx":"Apache-2.0"}},{"paper_slug":"random-erasing-data-augmentation","repo_url":"https://github.com/albumentations-team/albumentations","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"pytorch","reach":null},{"paper_slug":"random-erasing-data-augmentation","repo_url":"https://github.com/layumi/Person_reID_baseline_pytorch","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"pytorch","reach":null},{"paper_slug":"random-erasing-data-augmentation","repo_url":"https://github.com/layumi/University1652-Baseline","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"random-erasing-data-augmentation","repo_url":"https://github.com/pytorch/vision","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"data-augmentation","task_name":"Data Augmentation"},{"task_slug":"classification","task_name":"General Classification"},{"task_slug":"image-augmentation","task_name":"Image Augmentation"},{"task_slug":"image-classification","task_name":"Image Classification"},{"task_slug":"object-detection","task_name":"Object Detection"},{"task_slug":"person-re-identification","task_name":"Person Re-Identification"},{"task_slug":"robust-object-detection","task_name":"Robust Object Detection"},{"task_slug":"image-classification","task_name":"image-classification"},{"task_slug":"object-detection-1","task_name":"object-detection"}],"methods":[{"method_slug":"random-erasing","method_name":"Random Erasing"}],"datasets_introduced":[],"methods_introduced":[{"slug":"random-erasing","name":"Random Erasing","full_name":"Random Erasing"}],"results":[{"leaderboard":"/sota/image-classification-on-fashion-mnist","task":"Image Classification","dataset":"Fashion-MNIST","model":"Random Erasing","rank_in_archive_order":2,"of":34,"metrics":{"Percentage error":"3.65"},"uses_additional_data":false},{"leaderboard":"/sota/object-detection-on-pascal-voc-2007","task":"Object Detection","dataset":"PASCAL VOC 2007","model":"I+ORE","rank_in_archive_order":18,"of":30,"metrics":{"MAP":"76.2%"},"uses_additional_data":false},{"leaderboard":"/sota/person-re-identification-on-dukemtmc-reid","task":"Person Re-Identification","dataset":"DukeMTMC-reID","model":"SVDNet + Random Erasing","rank_in_archive_order":72,"of":94,"metrics":{"Rank-1":"79.3","mAP":"62.4"},"uses_additional_data":false},{"leaderboard":"/sota/person-re-identification-on-dukemtmc-reid","task":"Person Re-Identification","dataset":"DukeMTMC-reID","model":"TriNet + Random Erasing","rank_in_archive_order":77,"of":94,"metrics":{"Rank-1":"73.0","mAP":"56.6"},"uses_additional_data":false},{"leaderboard":"/sota/robust-object-detection-on-cityscapes-1","task":"Robust Object Detection","dataset":"Cityscapes","model":"Cutout","rank_in_archive_order":12,"of":13,"metrics":{"mPC [AP]":"15.7"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1708.04896","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}