Redefining Generalization in Visual Domains: A Two-Axis Framework for Fake Image Detection with FusionDetect

2025-10-07

AI Summary

To address insufficient cross-generator and cross-visual-domain generalization in synthetic image detection, this paper formally defines the dual generalization challenge within visual domains for the first time. We introduce OmniGen—a novel, comprehensive benchmark covering multiple generative models, diverse visual scenes, and common image perturbations—constituting the first evaluation platform of its kind. Furthermore, we propose FusionDetect, a lightweight yet effective detector that freezes CLIP and DINOv2 to extract complementary semantic and textural features, fuses them into a compact joint embedding space, and leverages multi-source synthetic data alongside perturbation-robust training. Extensive experiments demonstrate consistent improvements: +3.87% accuracy and +6.13% precision on mainstream benchmarks, and +4.48% accuracy on OmniGen, significantly enhancing both robustness and generalizability across generators and visual domains.

Technology Category

• Computer Vision: Generative Adversarial Networks (GANs) for Vision
• Natural Language Processing: Generation
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods

Abstract

The rapid development of generative models has made it increasingly crucial to develop detectors that can reliably detect synthetic images. Although most of the work has now focused on cross-generator generalization, we argue that this viewpoint is too limited. Detecting synthetic images involves another equally important challenge: generalization across visual domains. To bridge this gap,we present the OmniGen Benchmark. This comprehensive evaluation dataset incorporates 12 state-of-the-art generators, providing a more realistic way of evaluating detector performance under realistic conditions. In addition, we introduce a new method, FusionDetect, aimed at addressing both vectors of generalization. FusionDetect draws on the benefits of two frozen foundation models: CLIP & Dinov2. By deriving features from both complementary models,we develop a cohesive feature space that naturally adapts to changes in both thecontent and design of the generator. Our extensive experiments demonstrate that FusionDetect delivers not only a new state-of-the-art, which is 3.87% more accurate than its closest competitor and 6.13% more precise on average on established benchmarks, but also achieves a 4.48% increase in accuracy on OmniGen,along with exceptional robustness to common image perturbations. We introduce not only a top-performing detector, but also a new benchmark and framework for furthering universal AI image detection. The code and dataset are available at http://github.com/amir-aman/FusionDetect

Problem

• Detecting synthetic images across diverse visual domains
• Addressing cross-generator and cross-domain generalization challenges
• Developing robust fake image detection with complementary foundation models

Innovation

• FusionDetect uses CLIP and DINOv2 foundation models
• It creates a cohesive feature space for generalization
• Method addresses both generator and domain generalization