Adversarial Noise Layer: Regularize Neural Network By Adding Noise
This addresses regularization for CNNs to enhance model robustness, though it appears incremental as it builds on existing noise-based methods.
The paper tackles overfitting in CNNs by introducing Adversarial Noise Layer (ANL) and Class Adversarial Noise Layer (CANL), which add crafted noise to intermediate activations, resulting in improved generalization and robustness to adversarial examples like those from FGSM.
In this paper, we introduce a novel regularization method called Adversarial Noise Layer (ANL) and its efficient version called Class Adversarial Noise Layer (CANL), which are able to significantly improve CNN's generalization ability by adding carefully crafted noise into the intermediate layer activations. ANL and CANL can be easily implemented and integrated with most of the mainstream CNN-based models. We compared the effects of the different types of noise and visually demonstrate that our proposed adversarial noise instruct CNN models to learn to extract cleaner feature maps, which further reduce the risk of over-fitting. We also conclude that models trained with ANL or CANL are more robust to the adversarial examples generated by FGSM than the traditional adversarial training approaches.