Anti-Bandit Neural Architecture Search for Model Defense

Abstract

Deep convolutional neural networks (DCNNs) have dominated as the best performers in machine learning, but can be challenged by adversarial attacks. In this paper, we defend against adversarial attacks using neural architecture search (NAS) which is based on a comprehensive search of denoising blocks, weight-free operations, Gabor filters and convolutions. The resulting anti-bandit NAS (ABanditNAS) incorporates a new operation evaluation measure and search process based on the lower and upper confidence bounds (LCB and UCB). Unlike the conventional bandit algorithm using UCB for evaluation only, we use UCB to abandon arms for search efficiency and LCB for a fair competition between arms. Extensive experiments demonstrate that ABanditNAS is faster than other NAS methods, while achieving an $8.73\%$ improvement over prior arts on CIFAR-10 under PGD-$7$.

References in corpus (7)

• Improved Regularization of Convolutional Neural Networks with Cutout
• DARTS: Differentiable Architecture Search
• Delving into Transferable Adversarial Examples and Black-box Attacks
• Fast is better than free: Revisiting adversarial training
• A study of the effect of JPG compression on adversarial images
• Keeping the Bad Guys Out: Protecting and Vaccinating Deep Learning with JPEG Compression
• Evolving Robust Neural Architectures to Defend from Adversarial Attacks

Cited by in corpus (1)

• Smooth Adversarial Training