mlsanityData to 7 Oct 2026

On Interaction Between Augmentations and Corruptions in Natural Corruption Robustness

arXiv 2102.11273 ↗First posted 22 Feb 2021cs.CV cs.LGcited by 98

Not on the map. Why?

Abstract (arXiv)

Invariance to a broad array of image corruptions, such as warping, noise, or color shifts, is an important aspect of building robust models in computer vision. Recently, several new data augmentations have been proposed that significantly improve performance on ImageNet-C, a benchmark of such corruptions. However, there is still a lack of basic understanding on the relationship between data augmentations and test-time corruptions. To this end, we develop a feature space for image transforms, and then use a new measure in this space between augmentations and corruptions called the Minimal Sample Distance to demonstrate a strong correlation between similarity and performance. We then investigate recent data augmentations and observe a significant degradation in corruption robustness when the test-time corruptions are sampled to be perceptually dissimilar from ImageNet-C in this feature space. Our results suggest that test error can be improved by training on perceptually similar augmentations, and data augmentations may not generalize well beyond the existing benchmark. We hope our results and tools will allow for more robust progress towards improving robustness to image corruptions. We provide code at https://github.com/facebookresearch/augmentation-corruption.

Key facts. Cited by 98 papers in the map (data through 7 Oct 2026). Builds on 34 of its 48 references that are in the map. First posted on arXiv 22 Feb 2021.

Builds on

34 of 48 references.

References · 34

  1. 2015Deep Residual Learning for Image Recognitioncited by 51,809
  2. 2020An Image is Worth 16x16 Words: Transformers for Image Recognition at Scalecited by 29,915
  3. 2014Very Deep Convolutional Networks for Large-Scale Image Recognitioncited by 15,267
  4. 2018The Unreasonable Effectiveness of Deep Features as a Perceptual Metriccited by 11,029
  5. 2019EfficientNet: Rethinking Model Scaling for Convolutional Neural Networkscited by 6,415
  6. 2017mixup: Beyond Empirical Risk Minimizationcited by 5,155
  7. 2013Intriguing properties of neural networkscited by 5,114
  8. 2020Training data-efficient image transformers & distillation through attentioncited by 4,560
  9. 2016Wide Residual Networkscited by 3,057
  10. 2019Benchmarking Neural Network Robustness to Common Corruptions and Perturbationscited by 2,857
  11. 2020The Many Faces of Robustness: A Critical Analysis of Out-of-Distribution Generalizationcited by 1,842
  12. 2019Do ImageNet Classifiers Generalize to ImageNet?cited by 1,507
22 more
  1. 2017Accurate, Large Minibatch SGD: Training ImageNet in 1 Hourcited by 1,479
  2. 2018ImageNet-trained CNNs are biased towards texture; increasing shape bias improves accuracy and robustnesscited by 1,430
  3. 2019Self-training with Noisy Student improves ImageNet classificationcited by 1,304
  4. 2020Designing Network Design Spacescited by 954
  5. 2019AugMix: A Simple Data Processing Method to Improve Robustness and Uncertaintycited by 897
  6. 2020Do We Really Need to Access the Source Data? Source Hypothesis Transfer for Unsupervised Domain Adaptationcited by 844
  7. 2020Image Quality Assessment: Unifying Structure and Texture Similaritycited by 792
  8. 2018Exploring the Limits of Weakly Supervised Pretrainingcited by 604
  9. 2020Measuring Robustness to Natural Distribution Shifts in Image Classificationcited by 452
  10. 2020ResNeSt: Split-Attention Networkscited by 445
  11. 2020Improving robustness against common corruptions by covariate shift adaptationcited by 427
  12. 2019A Fourier Perspective on Model Robustness in Computer Visioncited by 288
  13. 2018Generalisation in humans and deep neural networkscited by 237
  14. 2018Do CIFAR-10 Classifiers Generalize to CIFAR-10?cited by 197
  15. 2017A Study and Comparison of Human and Deep Learning Recognition Performance Under Visual Distortionscited by 107
  16. 2018A Kernel Theory of Modern Data Augmentationcited by 106
  17. 2019Improving Robustness Without Sacrificing Accuracy with Patch Gaussian Augmentationcited by 105
  18. 2019On Network Design Spaces for Visual Recognitioncited by 68
  19. 2019Do Image Classifiers Generalize Across Time?cited by 59
  20. 2020On the Generalization Effects of Linear Transformations in Data Augmentationcited by 47
  21. 2020Compounding the Performance Improvements of Assembled Techniques in a Convolutional Neural Networkcited by 24
  22. 2019Robustness properties of Facebook's ResNeXt WSL modelscited by 17

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 7; 2022: 11; 2023: 25; 2024: 25; 2025: 18; 2026: 12.

98 papers.

Sort:

By area:

  1. 2022A ConvNet for the 2020scited by 3,545
  2. 2021PixMix: Dreamlike Pictures Comprehensively Improve Safety Measurescited by 131
  3. 20223D Common Corruptions and Data Augmentationcited by 120
  4. 2022Robust Mean Teacher for Continual and Gradual Test-Time Adaptationcited by 102
  5. 2025DriveSuprim: Towards Precise Trajectory Selection for End-to-End Planningcited by 91
  6. 2022Benchmarking Robustness of 3D Point Cloud Recognition Against Common Corruptionscited by 78
  7. 2024Wavelet Convolutions for Large Receptive Fieldscited by 74
  8. 2021On the effectiveness of adversarial training against common corruptionscited by 71
  9. 2022Back to the Source: Diffusion-Driven Test-Time Adaptationcited by 71
  10. 2021Test-Time Adaptation to Distribution Shift by Confidence Maximization and Input Transformationcited by 67
  11. 2023Universal Test-time Adaptation through Weight Ensembling, Diversity Weighting, and Prior Correctioncited by 65
  12. 2022ImageNet-X: Understanding Model Mistakes with Factor of Variation Annotationscited by 44
38 more
  1. 2024A Closer Look at the Robustness of Contrastive Language-Image Pre-Training (CLIP)cited by 44
  2. 2021PRIME: A few primitives can boost robustness to common corruptionscited by 40
  3. 2023Does CLIP's Generalization Performance Mainly Stem from High Train-Test Similarity?cited by 35
  4. 2021SimROD: A Simple Adaptation Method for Robust Object Detectioncited by 27
  5. 2023Large Separable Kernel Attention: Rethinking the Large Kernel Attention Design in CNNcited by 25
  6. 2024The Entropy Enigma: Success and Failure of Entropy Minimizationcited by 24
  7. 2022On the Strong Correlation Between Model Invariance and Generalizationcited by 23
  8. 2023Confidence and Dispersity Speak: Characterising Prediction Matrix for Unsupervised Accuracy Estimationcited by 22
  9. 2024Protected Test-Time Adaptation via Online Entropy Matching: A Betting Approachcited by 22
  10. 2021Dilated convolution with learnable spacingscited by 20
  11. 2023A Closer Look at Model Adaptation using Feature Distortion and Simplicity Biascited by 20
  12. 2023A Survey on the Robustness of Computer Vision Models against Common Corruptionscited by 20
  13. 2024Can Biases in ImageNet Models Explain Generalization?cited by 20
  14. 2024Fourier-basis Functions to Bridge Augmentation Gap: Rethinking Frequency Augmentation in Image Classificationcited by 18
  15. 2021Certified Adversarial Defenses Meet Out-of-Distribution Corruptions: Benchmarking Robustness and Simple Baselinescited by 17
  16. 2023Ethical Considerations for Responsible Data Curationcited by 17
  17. 2023Rethinking Precision of Pseudo Label: Test-Time Adaptation via Complementary Learningcited by 14
  18. 2023IPMix: Label-Preserving Data Augmentation Method for Training Robust Classifierscited by 14
  19. 2024HiKER-SGG: Hierarchical Knowledge Enhanced Robust Scene Graph Generationcited by 14
  20. 2025Out-of-Distribution Detection Methods Answer the Wrong Questionscited by 14
  21. 2024Toward a Holistic Evaluation of Robustness in CLIP Modelscited by 13
  22. 2022Introducing Intermediate Domains for Effective Self-Training during Test-Timecited by 12
  23. 2023HybridAugment++: Unified Frequency Spectra Perturbations for Model Robustnesscited by 12
  24. 2023CIFAR-10-Warehouse: Broad and More Realistic Testbeds in Model Generalization Analysiscited by 12
  25. 2023A Bag-of-Prototypes Representation for Dataset-Level Applicationscited by 11
  26. 2024Dynamic Sparse Training versus Dense Training: The Unexpected Winner in Image Corruption Robustnesscited by 11
  27. 2023It Is All About Data: A Survey on the Effects of Data on Adversarial Robustnesscited by 9
  28. 2023A Holistic Assessment of the Reliability of Machine Learning Systemscited by 9
  29. 2023Visual Data-Type Understanding does not emerge from Scaling Vision-Language Modelscited by 9
  30. 2023Benchmarking Robustness to Adversarial Image Obfuscationscited by 8
  31. 2024MANO: Exploiting Matrix Norm for Unsupervised Accuracy Estimation Under Distribution Shiftscited by 8
  32. 2024Improving robustness to corruptions with multiplicative weight perturbationscited by 7
  33. 2025Do ImageNet-trained models learn shortcuts? The impact of frequency shortcuts on generalizationcited by 5
  34. 2022Dynamic Test-Time Augmentation via Differentiable Functionscited by 4
  35. 2023Investigating the Corruption Robustness of Image Classifiers with Random Lp-norm Corruptionscited by 4
  36. 2024Phase-driven Domain Generalizable Learning for Nonstationary Time Seriescited by 4
  37. 2024Efficient Diffusion-Driven Corruption Editor for Test-Time Adaptationcited by 4
  38. 2023Understanding Data Augmentation from a Robustness Perspectivecited by 3

50 most cited of 98 All 98 (JSON)