{"url":"/method/autoaugment","slug":"autoaugment","name":"AutoAugment","full_name":"AutoAugment","full_name_withheld":false,"description_markdown":"**AutoAugment** is an automated approach to find data augmentation policies from data. It formulates the problem of finding the best augmentation policy as a discrete search problem. It consists of two components: a search algorithm and a search space. \r\n\r\nAt a high level, the search algorithm (implemented as a controller RNN) samples a data augmentation policy $S$, which has information about what image processing operation to use, the probability of using the operation in each batch, and the magnitude of the operation. The policy $S$ is used to train a neural network with a fixed architecture, whose validation accuracy $R$ is sent back to update the controller. Since $R$ is not differentiable, the controller will be updated by policy gradient methods. \r\n\r\nThe operations used are from PIL, a popular Python image library: all functions in PIL that accept an image as input and output an image. It additionally uses two other augmentation techniques: [Cutout](https://paperswithcode.com/method/cutout) and SamplePairing. The operations searched over are ShearX/Y, TranslateX/Y, Rotate, AutoContrast, Invert, Equalize, Solarize, Posterize, Contrast, Color, Brightness, Sharpness, Cutout and Sample Pairing.","description_state":"present","introduced_year":null,"introduced_by":{"title":null,"paper":null,"first_author":null,"n_authors":0,"url_abs":null,"archive_paper_url":null},"source":{"url":"http://arxiv.org/abs/1805.09501v3","title":"AutoAugment: Learning Augmentation Policies from Data","url_on_a_paper_host":true},"code_snippet_url":null,"code_snippet_url_on_a_code_host":false,"categories":[{"area":"Computer Vision","area_id":"computer-vision","collection":"Image Data Augmentation","url":"/methods/category/image-data-augmentation","pwc_aliases":[]}],"n_papers_tagged":60,"archive_num_papers":null,"papers_newest_first":[{"paper":null,"title":"Data Augmentation For Small Object using Fast AutoAugment","date":"2025-06-10","arxiv_id":"2506.08956","n_code_links":0,"syntology":null},{"paper":"/paper/learning-tree-structured-composition-of-data","title":"Learning Tree-Structured Composition of Data Augmentation","date":"2024-08-26","arxiv_id":"2408.14381","n_code_links":1,"syntology":null},{"paper":"/paper/segstrong-c-segmenting-surgical-tools","title":"SegSTRONG-C: Segmenting Surgical Tools Robustly On Non-adversarial Generated Corruptions -- An EndoVis'24 Challenge","date":"2024-07-16","arxiv_id":"2407.11906","n_code_links":1,"syntology":null},{"paper":"/paper/autoaugment-is-what-you-need-enhancing-rule","title":"AutoAugment Is What You Need: Enhancing Rule-based Augmentation Methods in Low-resource Regimes","date":"2024-02-08","arxiv_id":"2402.05584","n_code_links":1,"syntology":null},{"paper":null,"title":"AutoAugment Input Transformation for Highly Transferable Targeted Attacks","date":"2023-12-21","arxiv_id":"2312.14218","n_code_links":0,"syntology":null},{"paper":"/paper/la3-efficient-label-aware-autoaugment","title":"LA3: Efficient Label-Aware AutoAugment","date":"2023-04-20","arxiv_id":"2304.10310","n_code_links":1,"syntology":null},{"paper":"/paper/rangeaugment-efficient-online-augmentation","title":"RangeAugment: Efficient Online Augmentation with Range Learning","date":"2022-12-20","arxiv_id":"2212.10553","n_code_links":1,"syntology":null},{"paper":"/paper/efficient-and-effective-augmentation-strategy","title":"Efficient and Effective Augmentation Strategy for Adversarial Training","date":"2022-10-27","arxiv_id":"2210.15318","n_code_links":2,"syntology":{"ran":2,"of":7,"unverified":5,"pointer_only":0}},{"paper":"/paper/learning-self-regularized-adversarial-views","title":"Learning Self-Regularized Adversarial Views for Self-Supervised Vision Transformers","date":"2022-10-16","arxiv_id":"2210.08458","n_code_links":1,"syntology":null},{"paper":"/paper/ultra-fast-image-categorization-in-vivo-and","title":"Ultrafast Image Categorization in Biology and Neural Models","date":"2022-05-07","arxiv_id":"2205.03635","n_code_links":1,"syntology":null},{"paper":"/paper/deep-autoaugment-1","title":"Deep AutoAugment","date":"2022-03-11","arxiv_id":"2203.06172","n_code_links":1,"syntology":null},{"paper":null,"title":"SelectAugment: Hierarchical Deterministic Sample Selection for Data Augmentation","date":"2021-12-06","arxiv_id":"2112.02862","n_code_links":0,"syntology":null},{"paper":null,"title":"Challenges of Adversarial Image Augmentations","date":"2021-11-24","arxiv_id":"2111.12427","n_code_links":0,"syntology":null},{"paper":null,"title":"Learning Augmentation Distributions using Transformed Risk Minimization","date":"2021-11-16","arxiv_id":"2111.08190","n_code_links":0,"syntology":null},{"paper":"/paper/improving-model-generalization-by-agreement","title":"Improving Model Generalization by Agreement of Learned Representations from Data Augmentation","date":"2021-10-20","arxiv_id":"2110.10536","n_code_links":1,"syntology":null},{"paper":"/paper/text-autoaugment-learning-compositional","title":"Text AutoAugment: Learning Compositional Augmentation Policy for Text Classification","date":"2021-09-01","arxiv_id":"2109.00523","n_code_links":1,"syntology":null},{"paper":"/paper/direct-differentiable-augmentation-search","title":"Direct Differentiable Augmentation Search","date":"2021-04-09","arxiv_id":"2104.04282","n_code_links":1,"syntology":null},{"paper":null,"title":"Relating Adversarially Robust Generalization to Flat Minima","date":"2021-04-09","arxiv_id":"2104.04448","n_code_links":0,"syntology":null},{"paper":"/paper/robust-and-accurate-object-detection-via","title":"Robust and Accurate Object Detection via Adversarial Learning","date":"2021-03-23","arxiv_id":"2103.13886","n_code_links":3,"syntology":null},{"paper":"/paper/patch-autoaugment","title":"Local Patch AutoAugment with Multi-Agent Collaboration","date":"2021-03-20","arxiv_id":"2103.11099","n_code_links":2,"syntology":null},{"paper":"/paper/autodo-robust-autoaugment-for-biased-data","title":"AutoDO: Robust AutoAugment for Biased Data with Label Noise via Scalable Probabilistic Implicit Differentiation","date":"2021-03-10","arxiv_id":"2103.05863","n_code_links":1,"syntology":{"ran":5,"of":7,"unverified":2,"pointer_only":7}},{"paper":null,"title":"Faster and Smarter AutoAugment: Augmentation Policy Search Based on Dynamic Data-Clustering","date":"2021-01-01","arxiv_id":null,"n_code_links":0,"syntology":null},{"paper":null,"title":"NOSE Augment: Fast and Effective Data Augmentation Without Searching","date":"2021-01-01","arxiv_id":null,"n_code_links":0,"syntology":null},{"paper":null,"title":"Semantic Aware Data Augmentation for Cell Nuclei Microscopical Images With Artificial Neural Networks","date":"2021-01-01","arxiv_id":null,"n_code_links":0,"syntology":null},{"paper":null,"title":"StackMix: A complementary Mix algorithm","date":"2020-11-25","arxiv_id":"2011.12618","n_code_links":0,"syntology":null},{"paper":"/paper/keepaugment-a-simple-information-preserving","title":"KeepAugment: A Simple Information-Preserving Data Augmentation Approach","date":"2020-11-23","arxiv_id":"2011.11778","n_code_links":1,"syntology":null},{"paper":"/paper/an-efficient-and-scalable-deep-learning","title":"An Efficient and Scalable Deep Learning Approach for Road Damage Detection","date":"2020-11-18","arxiv_id":"2011.09577","n_code_links":2,"syntology":null},{"paper":"/paper/regularizing-neural-networks-via-adversarial","title":"Regularizing Neural Networks via Adversarial Model Perturbation","date":"2020-10-10","arxiv_id":"2010.04925","n_code_links":1,"syntology":{"ran":0,"of":12,"unverified":12,"pointer_only":0}},{"paper":null,"title":"A Technical Report for VIPriors Image Classification Challenge","date":"2020-07-17","arxiv_id":"2007.08722","n_code_links":0,"syntology":null},{"paper":null,"title":"Hypernetwork-Based Augmentation","date":"2020-06-11","arxiv_id":"2006.06320","n_code_links":0,"syntology":null}],"papers_shown":30,"tasks":[{"task":"/task/data-augmentation","name":"Data Augmentation","papers":41},{"task":"/task/image-classification","name":"Image Classification","papers":25},{"task":"/task/object-detection","name":"Object Detection","papers":14},{"task":"/task/image-classification","name":"image-classification","papers":13},{"task":"/task/object-detection-1","name":"object-detection","papers":11},{"task":"/task/classification","name":"General Classification","papers":8},{"task":"/task/domain-generalization","name":"Domain Generalization","papers":6},{"task":null,"name":"GPU","papers":5},{"task":"/task/image-augmentation","name":"Image Augmentation","papers":5},{"task":"/task/semantic-segmentation","name":"Semantic Segmentation","papers":5},{"task":"/task/automl","name":"AutoML","papers":4},{"task":"/task/diversity","name":"Diversity","papers":4},{"task":"/task/fine-grained-image-classification","name":"Fine-Grained Image Classification","papers":4},{"task":"/task/classification-1","name":"Classification","papers":3},{"task":"/task/contrastive-learning","name":"Contrastive Learning","papers":3},{"task":"/task/image-retrieval","name":"Image Retrieval","papers":3},{"task":"/task/knowledge-distillation","name":"Knowledge Distillation","papers":3},{"task":"/task/object","name":"Object","papers":3},{"task":"/task/reinforcement-learning","name":"Reinforcement Learning","papers":3},{"task":"/task/retrieval","name":"Retrieval","papers":3}],"tasks_shown":20,"n_tasks":68,"usage_by_year":[{"year":"2018","papers":2},{"year":"2019","papers":16},{"year":"2020","papers":18},{"year":"2021","papers":13},{"year":"2022","papers":5},{"year":"2023","papers":2},{"year":"2024","papers":3},{"year":"2025","papers":1}],"row_source":"embedded","archive":{"source":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","archive_url":"https://paperswithcode.com/method/autoaugment"},"syntology_read_at":"2026-09-24T18:15:14+00:00"}