ML Security PapersUniShield: An Adaptive Multi-Agent Framework for Unified Forgery Image Detection and Localization
Qing Huang 1,2, Zhipei Xu 1, Xuanyu Zhang 1, Jian Zhang 1
Published on arXiv
2510.03161
Output Integrity Attack
OWASP ML Top 10 — ML09
Key Finding
UniShield achieves state-of-the-art performance across multiple forgery domains, surpassing both existing unified approaches and domain-specific detectors in detection and localization tasks
UniShield
Novel technique introduced
With the rapid advancements in image generation, synthetic images have become increasingly realistic, posing significant societal risks, such as misinformation and fraud. Forgery Image Detection and Localization (FIDL) thus emerges as essential for maintaining information integrity and societal security. Despite impressive performances by existing domain-specific detection methods, their practical applicability remains limited, primarily due to their narrow specialization, poor cross-domain generalization, and the absence of an integrated adaptive framework. To address these issues, we propose UniShield, the novel multi-agent-based unified system capable of detecting and localizing image forgeries across diverse domains, including image manipulation, document manipulation, DeepFake, and AI-generated images. UniShield innovatively integrates a perception agent with a detection agent. The perception agent intelligently analyzes image features to dynamically select suitable detection models, while the detection agent consolidates various expert detectors into a unified framework and generates interpretable reports. Extensive experiments show that UniShield achieves state-of-the-art results, surpassing both existing unified approaches and domain-specific detectors, highlighting its superior practicality, adaptiveness, and scalability.
Key Contributions
- UniShield: a multi-agent unified framework that covers image manipulation, document manipulation, DeepFake, and AI-generated image detection in a single system
- Perception agent that intelligently analyzes image features to dynamically select the most suitable expert detection model for a given input
- Detection agent that consolidates diverse expert detectors and generates structured, interpretable forensic reports with description, localization, and judgment basis
🛡️ Threat Analysis
UniShield is a novel detection architecture for AI-generated content and forged images (deepfakes, AI-generated images, image/document manipulation), which falls squarely under output integrity and AI-generated content detection. The primary contribution is a novel multi-agent forensic framework, not mere application of existing detectors to a new domain.