Post-Training Augmentation Invariance
For practitioners needing robust pretrained models, this method provides a lightweight, fine-tuning-free way to add augmentation invariance, though it is incremental over existing adapter-based approaches.
The paper introduces a framework for adding invariance to augmentations in pretrained networks without altering original behavior, using lightweight adapter networks. On STL10 with DINOv2 features, it boosts rotated-image classification accuracy from 71% to 94% and noise-invariant accuracy from 58% to 86%.
This work develops a framework for post-training augmentation invariance, in which our goal is to add invariance properties to a pretrained network without altering its behavior on the original, non-augmented input distribution. We define this notion precisely and additionally introduce augmented encoders, which are probabilistic encoders that formalize augmentation-based encoding processes and that serve as our fundamental object of study. We introduce two losses for augmented encoders, namely, Markov-Wasserstein minimization and Wasserstein correlation maximization, and we demonstrate empirically that both losses can be used to train lightweight, one-hidden-layer MLP adapter networks E_theta that, when appended to the latent space of a pretrained network F, do indeed lead to (approximate) post-training augmentation invariance. For example, on STL10 with F = DINOv2 features, the composite network C o E_theta o F, where C is a linear classifier and where E_theta is one of our proposed adapter networks, achieves 94% classification accuracy on arbitrarily rotated images, whereas a network of the form C o F without the adapter E_theta drops to 71% accuracy. Similarly, we can boost noise-invariant classification results from 58% up to 86%. Significantly, we obtain these results with no fine-tuning (the weights of F remain frozen throughout), and our methods introduce little corruption to the original features, since E_theta acts nearly isometrically on the non-augmented latent distribution. In contrast, we show that adapter networks trained with alternative candidate losses, specifically SimCLR and HSIC maximization, produce uncompetitive classification results and fundamentally corrupt the original latent space. Code available at: https://github.com/keenan-eikenberry/augmentation_invariance