Robustness to Corruption in Pre-Trained Bayesian Neural Networks

Abstract

We develop ShiftMatch, a new training-data-dependent likelihood for robustness to corruption in Bayesian neural networks (BNNs). ShiftMatch is inspired by the training-data-dependent “EmpCov” priors from Izmailov et al. (2021a), and efficiently matches test-time spatial correlations to those at training time. Critically, ShiftMatch is designed to leave the neural network’s training time likelihood unchanged, allowing it to use publicly available samples from pre-trained BNNs. Using pre-trained HMC samples, ShiftMatch gives strong performance improvements on CIFAR-10-C, outperforms EmpCov priors (though ShiftMatch uses extra information from a minibatch of corrupted test points), and is perhaps the first Bayesian method capable of convincingly outperforming plain deep ensembles.

Cite

Text

Wang and Aitchison. "Robustness to Corruption in Pre-Trained Bayesian Neural Networks." International Conference on Learning Representations, 2023.

Markdown

[Wang and Aitchison. "Robustness to Corruption in Pre-Trained Bayesian Neural Networks." International Conference on Learning Representations, 2023.](https://mlanthology.org/iclr/2023/wang2023iclr-robustness/)

BibTeX

@inproceedings{wang2023iclr-robustness,
  title     = {{Robustness to Corruption in Pre-Trained Bayesian Neural Networks}},
  author    = {Wang, Xi and Aitchison, Laurence},
  booktitle = {International Conference on Learning Representations},
  year      = {2023},
  url       = {https://mlanthology.org/iclr/2023/wang2023iclr-robustness/}
}