STD-FD: Spatio-Temporal Distribution Fitting Deviation for AIGC Forgery Identification

With the rise of AIGC technologies, particularly diffusion models, highly realistic fake images that can deceive human visual perception has become feasible. Consequently, various forgery detection methods have emerged. However, existing methods treat the generation process of fake images as either a black-box or an auxiliary tool, offering limited insights into its underlying mechanisms. In this paper, we propose Spatio-Temporal Distribution Fitting Deviation (STD-FD) for AIGC forgery detection, which explores the generative process in detail. By decomposing and reconstructing data within generative diffusion models, initial experiments reveal temporal distribution fitting deviations during the image reconstruction process. These deviations are captured through reconstruction noise maps for each spatial semantic unit, derived via a super-resolution algorithm. Critical discriminative patterns, termed DFactors, are identified through statistical modeling of these deviations. Extensive experiments show that STD-FD effectively captures distribution patterns in AIGC-generated data, demonstrating strong robustness and generalizability while outperforming state-of-the-art (SOTA) methods on major datasets. The source code is available at this link.

Fake photos and artwork made by today’s AI image generators can look so real that even experts struggle to spot them. We asked a simple question: When an AI “paints” a picture step-by-step, does it leave behind hidden fingerprints that a human eye can’t see?To find out, we zoomed in on the tiny changes the generator makes while refining an image. By replaying this process in slow motion, we discovered subtle, time-based “wobbles” in how colours and textures settle across different parts of the picture. We captured these wobbles as noise maps, then turned the strongest patterns—called DFactors—into tell-tale signs of fakery.Our detector, named STD-FD, reliably flags AI-generated images across several public test sets and beats today’s best forensic tools. This technique could help journalists, social-media platforms, and law-enforcement agencies spot deepfakes quickly, protecting the public from misinformation and visual fraud.