Beyond Artifacts: Real-Centric Envelope Modeling for Reliable AI-Generated Image Detection
This addresses the need for reliable detection of synthetic images in real-world scenarios, such as cross-platform sharing and post-processing, which is an incremental advancement over existing methods that overfit to artifacts.
The paper tackles the problem of AI-generated image detection under real-world conditions by proposing Real-centric Envelope Modeling (REM), which shifts from learning generator artifacts to modeling the robust distribution of real images, achieving an average improvement of 7.5% over state-of-the-art methods across eight benchmarks.
The rapid progress of generative models has intensified the need for reliable and robust detection under real-world conditions. However, existing detectors often overfit to generator-specific artifacts and remain highly sensitive to real-world degradations. As generative architectures evolve and images undergo multi-round cross-platform sharing and post-processing (chain degradations), these artifact cues become obsolete and harder to detect. To address this, we propose Real-centric Envelope Modeling (REM), a new paradigm that shifts detection from learning generator artifacts to modeling the robust distribution of real images. REM introduces feature-level perturbations in self-reconstruction to generate near-real samples, and employs an envelope estimator with cross-domain consistency to learn a boundary enclosing the real image manifold. We further build RealChain, a comprehensive benchmark covering both open-source and commercial generators with simulated real-world degradation. Across eight benchmark evaluations, REM achieves an average improvement of 7.5% over state-of-the-art methods, and notably maintains exceptional generalization on the severely degraded RealChain benchmark, establishing a solid foundation for synthetic image detection under real-world conditions. The code and the RealChain benchmark will be made publicly available upon acceptance of the paper.