Robustness via Cross-Domain Ensembles
This work addresses robustness issues in neural networks for applications like computer vision, though it appears incremental as it builds on existing ensemble techniques.
The paper tackles the problem of neural network robustness to distribution shifts by introducing a method that ensembles predictions from diverse cues, achieving significantly greater robustness than standard learning, deep ensembles, and other baselines across multiple tasks and datasets.
We present a method for making neural network predictions robust to shifts from the training data distribution. The proposed method is based on making predictions via a diverse set of cues (called 'middle domains') and ensembling them into one strong prediction. The premise of the idea is that predictions made via different cues respond differently to a distribution shift, hence one should be able to merge them into one robust final prediction. We perform the merging in a straightforward but principled manner based on the uncertainty associated with each prediction. The evaluations are performed using multiple tasks and datasets (Taskonomy, Replica, ImageNet, CIFAR) under a wide range of adversarial and non-adversarial distribution shifts which demonstrate the proposed method is considerably more robust than its standard learning counterpart, conventional deep ensembles, and several other baselines.