Secure Video Quality Assessment Resisting Adversarial Attacks

2024-10-09

arXiv.org

AI Summary

Current video quality assessment (VQA) models exhibit high vulnerability to adversarial attacks, and no robust defense mechanisms have been established. To address this critical gap, we propose SecureVQA—the first security-aware framework for VQA—introducing a novel tripartite defense paradigm: (i) random spatial grid sampling, (ii) pixel-level guardian map randomization, and (iii) spatiotemporal collaborative defense. By integrating temporal feature compensation and adversarial robust training, SecureVQA preserves state-of-the-art (SOTA) evaluation accuracy while substantially enhancing robustness against both white-box and black-box attacks. Extensive experiments across multiple benchmarks demonstrate an average 42.6% reduction in attack success rate. SecureVQA establishes the first standardized benchmark for secure VQA evaluation and delivers a deployable defense infrastructure for trustworthy video quality analysis.

Technology Category

• Computer Vision: Adversarial Attacks & Robustness
• Machine Learning: Adversarial Learning & Robustness
• Multiagent Systems: Adversarial Agents

Application Category

• Security and Privacy: Large-scale security measurements
• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Search and Retrieval-Augmented AI: Vertical and domain-specific search

Abstract

The exponential surge in video traffic has intensified the imperative for Video Quality Assessment (VQA). Leveraging cutting-edge architectures, current VQA models have achieved human-comparable accuracy. However, recent studies have revealed the vulnerability of existing VQA models against adversarial attacks. To establish a reliable and practical assessment system, a secure VQA model capable of resisting such malicious attacks is urgently demanded. Unfortunately, no attempt has been made to explore this issue. This paper first attempts to investigate general adversarial defense principles, aiming at endowing existing VQA models with security. Specifically, we first introduce random spatial grid sampling on the video frame for intra-frame defense. Then, we design pixel-wise randomization through a guardian map, globally neutralizing adversarial perturbations. Meanwhile, we extract temporal information from the video sequence as compensation for inter-frame defense. Building upon these principles, we present a novel VQA framework from the security-oriented perspective, termed SecureVQA. Extensive experiments indicate that SecureVQA sets a new benchmark in security while achieving competitive VQA performance compared with state-of-the-art models. Ablation studies delve deeper into analyzing the principles of SecureVQA, demonstrating their generalization and contributions to the security of leading VQA models.

Problem

• Secure VQA resisting adversarial attacks
• Vulnerability of existing VQA models
• Defense principles for video quality assessment

Innovation

• Random spatial grid sampling for intra-frame defense
• Pixel-wise randomization via guardian map globally
• Temporal information extraction for inter-frame defense compensation