Corpus ID: 220250257

The Many Faces of Robustness: A Critical Analysis of Out-of-Distribution Generalization

@article{Hendrycks2020TheMF,
  title={The Many Faces of Robustness: A Critical Analysis of Out-of-Distribution Generalization},
  author={Dan Hendrycks and Steven Basart and Norman Mu and Saurav Kadavath and F. Wang and Evan Dorundo and Rahul Desai and Tyler Lixuan Zhu and Samyak Parajuli and M. Guo and D. Song and J. Steinhardt and J. Gilmer},
  journal={ArXiv},
  year={2020},
  volume={abs/2006.16241}
}
We introduce three new robustness benchmarks consisting of naturally occurring distribution changes in image style, geographic location, camera operation, and more. Using our benchmarks, we take stock of previously proposed hypotheses for out-of-distribution robustness and put them to the test. We find that using larger models and synthetic data augmentation can improve robustness on real-world distribution shifts, contrary to claims in prior work. Motivated by this, we introduce a new data… Expand
44 Citations
Revisiting Batch Normalization for Improving Corruption Robustness
  • 7
  • PDF
Improving robustness against common corruptions with frequency biased models
  • Highly Influenced
  • PDF
On Interaction Between Augmentations and Corruptions in Natural Corruption Robustness
  • Highly Influenced
  • PDF
Improved Handling of Motion Blur in Online Object Detection
  • PDF
Random Shadows and Highlights: A new data augmentation method for extreme lighting conditions
  • 1
  • PDF
Shape-Texture Debiased Neural Network Training
  • 2
  • Highly Influenced
  • PDF
Natural Perturbed Training for General Robustness of Neural Network Classifiers
  • PDF
...
1
2
3
4
5
...

References

SHOWING 1-10 OF 44 REFERENCES
AugMix: A Simple Data Processing Method to Improve Robustness and Uncertainty
  • 140
  • PDF
Do ImageNet Classifiers Generalize to ImageNet?
  • 226
  • PDF
Exploring the Limits of Weakly Supervised Pretraining
  • 496
  • PDF
AutoAugment: Learning Augmentation Policies from Data
  • 573
  • PDF
Aggregated Residual Transformations for Deep Neural Networks
  • 3,418
  • PDF
CutMix: Regularization Strategy to Train Strong Classifiers With Localizable Features
  • 374
  • PDF
Learning Robust Global Representations by Penalizing Local Predictive Power
  • 41
  • PDF
The Unreasonable Effectiveness of Deep Features as a Perceptual Metric
  • 1,283
  • PDF
...
1
2
3
4
5
...