UniShield：面向统一伪造图像检测与定位的自适应多智能体框架

View PDF HTML (experimental)

Abstract: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.

Subjects:	Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
Cite as:	arXiv:2510.03161 [cs.CV]
	(or arXiv:2510.03161v4 [cs.CV] for this version)
	https://doi.org/10.48550/arXiv.2510.03161 arXiv-issued DOI via DataCite

Submission history

From: Qing Huang [view email]
[v1] Fri, 3 Oct 2025 16:33:05 UTC (1,898 KB)
[v2] Fri, 15 May 2026 16:00:57 UTC (1,730 KB)
[v3] Wed, 23 Sep 2026 07:41:00 UTC (1,730 KB)
[v4] Thu, 8 Oct 2026 08:08:10 UTC (1,730 KB)