A Survey and Evaluation of Adversarial Attacks for Object Detection

2024-08-04

arXiv.org

AI Summary

Object detection models exhibit significant vulnerability to adversarial attacks, yet robustness evaluation lacks standardization and comprehensive theoretical frameworks. Method: We propose the first adversarial attack taxonomy specifically for object detection, systematically unifying and extending robustness evaluation metrics. Through empirical analysis of 12 representative attacks—including FGSM, PGD, and CIoU-loss—across diverse detectors (e.g., Faster R-CNN, YOLOv5, and vision-language pretrained models like Grounding DINO), we characterize cross-architecture attack failure patterns and critical vulnerability modes. Contribution/Results: We identify severe inconsistencies in current evaluation practices and formally establish the necessity of a unified benchmarking protocol. Key research gaps are pinpointed, including multi-scale perturbation modeling and joint optimization of bounding-box confidence scores and IoU. Our framework provides both theoretical foundations and practical guidelines for developing robust detection systems.

Technology Category

• Computer Vision: Adversarial Attacks & Robustness
• Machine Learning: Adversarial Learning & Robustness
• Search and Optimization: Adversarial Search

Application Category

• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Security and Privacy: Large-scale security measurements

Abstract

Deep learning models achieve remarkable accuracy in computer vision tasks, yet remain vulnerable to adversarial examples--carefully crafted perturbations to input images that can deceive these models into making confident but incorrect predictions. This vulnerability pose significant risks in high-stakes applications such as autonomous vehicles, security surveillance, and safety-critical inspection systems. While the existing literature extensively covers adversarial attacks in image classification, comprehensive analyses of such attacks on object detection systems remain limited. This paper presents a novel taxonomic framework for categorizing adversarial attacks specific to object detection architectures, synthesizes existing robustness metrics, and provides a comprehensive empirical evaluation of state-of-the-art attack methodologies on popular object detection models, including both traditional detectors and modern detectors with vision-language pretraining. Through rigorous analysis of open-source attack implementations and their effectiveness across diverse detection architectures, we derive key insights into attack characteristics. Furthermore, we delineate critical research gaps and emerging challenges to guide future investigations in securing object detection systems against adversarial threats. Our findings establish a foundation for developing more robust detection models while highlighting the urgent need for standardized evaluation protocols in this rapidly evolving domain.

Problem

• Analyzes adversarial attacks on object detection systems
• Evaluates attack effectiveness across diverse detection architectures
• Identifies research gaps for robust object detection models

Innovation

• Novel taxonomic framework for adversarial attacks
• Comprehensive empirical evaluation of attack methodologies
• Standardized evaluation protocols for robustness