Robustness properties of Facebook's ResNeXt WSL models

Do ImageNet classifiers generalize to ImageNet?

Original

Recht, B., Roelofs, R., Schmidt, L., and Shankar, V. (2019) · 1902

Earlier work this paper cites.

Learning perceptually-aligned representations via adversarial robustness

Original

Engstrom, L., Ilyas, A., Santurkar, S., Tsipras, D., Tran, B., and Madry, A. (2019) · 1906

Earlier work this paper cites.

Improving robustness without sacrificing accuracy with patch Gaussian augmentation

Original

Lopes, R. G., Yin, D., Poole, B., Gilmer, J., and Cubuk, E. D. (2019) · 1906

Earlier work this paper cites.

Natural adversarial examples

Original

Hendrycks, D., Zhao, K., Basart, S., Steinhardt, J., and Song, D. (2019) · 1907

Earlier work this paper cites.

The unreasonable effectiveness of data

Halevy, A., Norvig, P., and Pereira, F. (2009) · 2009

Earlier work this paper cites.

Foolbox: A Python toolbox to benchmark the robustness of machine learning models

Original

Rauber, J., Brendel, W., and Bethge, M. (2017) · 2017

Cited alongside, same era.

Aggregated residual transformations for deep neural networks

Xie, S., Girshick, R., Dollár, P., Tu, Z., and He, K. (2017) · 2017

Cited alongside, same era.

Evaluating and understanding the robustness of adversarial logit pairing

Original

Engstrom, L., Ilyas, A., and Athalye, A. (2018) · 2018

Cited alongside, same era.

Adversarial logit pairing

Original

Kannan, H., Kurakin, A., and Goodfellow, I. (2018) · 2018

Cited alongside, same era.

Exploring the limits of weakly supervised pretraining

Mahajan, D., Girshick, R., Ramanathan, V., He, K., Paluri, M., Li, Y., Bharambe, A., and van der Maaten, L. (2018) · 2018

Cited alongside, same era.

Feature denoising for improving adversarial robustness

Original

Xie, C., Wu, Y., van der Maaten, L., Yuille, A., and He, K. (2018) · 2018

Later among the works it cites.

Approximating CNNs with bag-of-local-features models works surprisingly well on ImageNet

Brendel, W. and Bethge, M. (2019) · 2019

Closest in time.

ImageNet-trained CNNs are biased toward texture; increasing shape bias improves accuracy and robustness

Geirhos, R., Rubisch, P., Michaelis, C., Bethge, M., Wichmann, F. A., and Brendel, W. (2019) · 2019

Closest in time.

Adversarial examples are a natural consequence of test error in noise

Gilmer, J., Ford, N., Carlini, N., and Cubuk, E. (2019) · 2019

Closest in time.

Benchmarking neural network robustness to common corruptions and perturbations

Hendrycks, D. and Dietterich, T. (2019) · 2019

Closest in time.