Overcomplete Representations Against Adversarial Videos

TL;DR

This paper introduces OUDefend, a novel over-and-under complete restoration network designed to improve the robustness of video recognition models against various adversarial attacks by balancing local and global feature representations.

Contribution

The paper proposes OUDefend, a new restoration network that combines overcomplete and undercomplete representations to defend against adversarial videos, which is a novel approach in this domain.

Findings

OUDefend improves robustness against additive, multiplicative, and physical adversarial video attacks.

The method effectively balances local and global features for better defense.

Experimental results demonstrate significant robustness gains over existing defenses.

Abstract

Adversarial robustness of deep neural networks is an extensively studied problem in the literature and various methods have been proposed to defend against adversarial images. However, only a handful of defense methods have been developed for defending against attacked videos. In this paper, we propose a novel Over-and-Under complete restoration network for Defending against adversarial videos (OUDefend). Most restoration networks adopt an encoder-decoder architecture that first shrinks spatial dimension then expands it back. This approach learns undercomplete representations, which have large receptive fields to collect global information but overlooks local details. On the other hand, overcomplete representations have opposite properties. Hence, OUDefend is designed to balance local and global features by learning those two representations. We attach OUDefend to target video recognition models as a feature restoration block and train the entire network end-to-end. Experimental results show that the defenses focusing on images may be ineffective to videos, while OUDefend enhances robustness against different types of adversarial videos, ranging from additive attacks, multiplicative attacks to physically realizable attacks. Code: https://github.com/shaoyuanlo/OUDefend