Rethinking the Use of Vision Transformers for AI-Generated Image Detection

Rich feature representations derived from CLIP-ViT have been widely utilized in AI-generated image detection. While most existing methods primarily leverage features from the final layer, we systematically analyze the contributions of layer-wise features to this task. Our study reveals that earlier layers provide more localized and generalizable features, often surpassing the performance of final-layer features in detection tasks. Moreover, we find that different layers capture distinct aspects of the data, each contributing uniquely to AI-generated image detection. Motivated by these findings, we introduce a novel adaptive method, termed MoLD, which dynamically integrates features from multiple ViT layers using a gating-based mechanism. Extensive experiments on both GAN- and diffusion-generated images demonstrate that MoLD significantly improves detection performance, enhances generalization across diverse generative models, and exhibits robustness in real-world scenarios. Finally, we illustrate the scalability and versatility of our approach by successfully applying it to other pre-trained ViTs, such as DINOv2.

Key Contributions

• Systematic layer-wise analysis of CLIP-ViT features showing earlier layers provide more localizable and generalizable representations for AI-generated image detection
• MoLD: a gating-based adaptive mechanism that dynamically fuses features across multiple ViT layers for improved detection performance
• Demonstrated scalability of MoLD to other pre-trained ViTs (DINOv2) and robustness across diverse GAN and diffusion model generators

🛡️ Threat Analysis

Details

Similar Papers