mlsanityData to 7 Oct 2026

Deep AutoAugment

arXiv 2203.06172 ↗First posted 11 Mar 2022cs.CV cs.AIcited by 18

Not on the map. Why?

Abstract (arXiv)

While recent automated data augmentation methods lead to state-of-the-art results, their design spaces and the derived data augmentation strategies still incorporate strong human priors. In this work, instead of fixing a set of hand-picked default augmentations alongside the searched data augmentations, we propose a fully automated approach for data augmentation search named Deep AutoAugment (DeepAA). DeepAA progressively builds a multi-layer data augmentation pipeline from scratch by stacking augmentation layers one at a time until reaching convergence. For each augmentation layer, the policy is optimized to maximize the cosine similarity between the gradients of the original and augmented data along the direction with low variance. Our experiments show that even without default augmentations, we can learn an augmentation policy that achieves strong performance with that of previous works. Extensive ablation studies show that the regularized gradient matching is an effective search method for data augmentation policies. Our code is available at: https://github.com/MSU-MLSys-Lab/DeepAA .

Key facts. Cited by 18 papers in the map (data through 7 Oct 2026). Builds on 27 of its 32 references that are in the map. First posted on arXiv 11 Mar 2022.

Builds on

27 of 32 references.

References · 27

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2014Adam: A Method for Stochastic Optimizationcited by 36,960
  3. 2017mixup: Beyond Empirical Risk Minimizationcited by 5,155
  4. 2014Going Deeper with Convolutionscited by 4,231
  5. 2016Wide Residual Networkscited by 3,057
  6. 2019CutMix: Regularization Strategy to Train Strong Classifiers with Localizable Featurescited by 2,875
  7. 2016Categorical Reparameterization with Gumbel-Softmaxcited by 2,708
  8. 2019RandAugment: Practical automated data augmentation with a reduced search spacecited by 2,261
  9. 2018DARTS: Differentiable Architecture Searchcited by 1,782
  10. 2016Understanding deep learning requires rethinking generalizationcited by 1,657
  11. 2017Improved Regularization of Convolutional Neural Networks with Cutoutcited by 1,623
  12. 2019AugMix: A Simple Data Processing Method to Improve Robustness and Uncertaintycited by 897
15 more
  1. 2021ResNet strikes back: An improved training procedure in timmcited by 365
  2. 2019Fast AutoAugmentcited by 243
  3. 2021TrivialAugment: Tuning-free Yet State-of-the-Art Data Augmentationcited by 225
  4. 2020Improve Unsupervised Domain Adaptation with Mixup Trainingcited by 185
  5. 2019Population Based Augmentation: Efficient Learning of Augmentation Policy Schedulescited by 153
  6. 2019Adversarial AutoAugmentcited by 111
  7. 2018Adapting Auxiliary Losses Using Gradient Similaritycited by 98
  8. 2018Data Augmentation by Pairing Samples for Images Classificationcited by 96
  9. 2019Faster AutoAugment: Learning Augmentation Strategies using Backpropagationcited by 96
  10. 2019MultiGrain: a unified image embedding for classes and instancescited by 85
  11. 2019Optimizing Data Usage via Differentiable Rewardscited by 48
  12. 2020UniformAugment: A Search-free Probabilistic Data Augmentation Approachcited by 38
  13. 2021Drawing Multiple Augmentation Samples Per Image During Training Efficiently Decreases Test Errorcited by 32
  14. 2021Direct Differentiable Augmentation Searchcited by 28
  15. 2021In-Loop Meta-Learning with Gradient-Alignment Rewardcited by 1

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2022: 4; 2023: 6; 2024: 5; 2025: 1; 2026: 2.

18 papers.

Sort:

  1. 2024SF(DA)2: Source-free Domain Adaptation Through the Lens of Data Augmentationcited by 19
  2. 2022Automatic Data Augmentation via Invariance-Constrained Learningcited by 17
  3. 2022Learning Instance-Specific Augmentations by Capturing Local Invariancescited by 15
  4. 2024Automated data processing and feature engineering for deep learning and big data applications: a surveycited by 14
  5. 2023When to Learn What: Model-Adaptive Data Augmentation Curriculumcited by 13
  6. 2023Understanding the Detrimental Class-level Effects of Data Augmentationcited by 12
  7. 2022RangeAugment: Efficient Online Augmentation with Range Learningcited by 8
  8. 2024Data Augmentation Policy Search for Long-Term Forecastingcited by 5
  9. 2023Data Augmentation for Neural NLPcited by 2
  10. 2023LatentAugment: Dynamically Optimized Latent Probabilities of Data Augmentationcited by 1
  11. 2023Integrating GAN and Texture Synthesis for Enhanced Road Damage Detectioncited by 1
  12. 2024Data augmentation with automated machine learning: approaches and performance comparison with classical data augmentation methodscited by 1
6 more
  1. 2025Optimizing Data Augmentation through Bayesian Model Selectioncited by 1
  2. 2026One Model, Many Behaviors: Training-Induced Effects on Out-of-Distribution Detectioncited by 1
  3. 2022Expeditious Saliency-guided Mix-up through Random Gradient Thresholdingcited by 0
  4. 2023Unsupervised augmentation optimization for few-shot medical image segmentationcited by 0
  5. 2024Boosting Semi-Supervised 2D Human Pose Estimation by Revisiting Data Augmentation and Consistency Trainingcited by 0
  6. 2026Ctrl-A: Control-Driven Online Data Augmentationcited by 0