Parametric Noise Injection: Trainable Randomness to Improve Deep Neural Network Robustness Against Adversarial Attack

By Adnan Siraj Rakin, Zhezhi He, Deliang Fan. 2018

328 citations recorded by the source corpus.

Recent developments in the field of Deep Learning have exposed the underlying vulnerability of Deep Neural Network (DNN) against adversarial examples. In image classification, an adversarial example is a carefully modified image that is visually imperceptible to the original image but can cause DNN model to misclassify it. Training the network with Gaussian noise is an effective technique to perform model regularization, thus improving model robustness against input variation. Inspired by this classical method, we explore to utilize the regularization characteristic of noise injection to improve DNN's robustness against adversarial attack. In this work, we propose Parametric-Noise-Injection (PNI) which involves trainable Gaussian noise injection at each layer on either activation or weights through solving the Min-Max optimization problem, embedded with adversarial training. These parameters are trained explicitly to achieve improved robustness. The extensive results show that our proposed PNI technique effectively improves the robustness against a variety of powerful white-box and black-box attacks such as PGD, C&W, FGSM, transferable attack, and ZOO attack. Last but not the least, PNI method improves both clean- and perturbed-data accuracy, in comparison to the state-of-the-art defense methods, which outperforms current unbroken PGD defense by 1.1% and 6.8% on clean- and perturbed- test data respectively, using ResNet-20 architecture.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Safety, Robustness & Interpretability

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Vision and Pattern Recognition

In Safety, Robustness & Interpretability

Citations in the corpus

Theoretical evidence for adversarial robustness through randomization: the case of the Exponential family (2019, background): “Several papers [26, 37] propose to inject Gaussian noise directly on the activation of selected layers both at training and inference time.”
Adversarial Weight Perturbation Helps Robust Generalization (2020, methodology): “Different from random perturbations [16], AWP injects the strongest worst-case weight perturbations, forming a double-perturbation mechanism (i.”
Bit-Flip Attack: Crushing Neural Network With Progressive Bit Search (2019, background): “Both attack and defense of adversarial example on the input end of DNN has been heavily investigated in the past couple of years [7, 6, 8] and still be in progress [9, 10, 11].”
Advances in adversarial attacks and defenses in computer vision: A survey (2021, background): “[376] proposed a trainable Gaussian model for injecting the noise.”
A Survey of Black-Box Adversarial Attacks on Computer Vision Models (2019, background): “Month Year Method Proposed Approach Mar 2016 Papernot et al. [46] Distillation Mar 2017 Hosseini et al. [47] NULL Labeling Sep 2017 Meng et al. [48] MagNet Dec 2017 Xu et al. [49] Feature Squeezing Jan 2018 Guo et al. [50] Input Transformations Feb 2018 Buckman et al. [51] Thermometer Encoding Feb 2”