Anchor-Regularized Adaptation for Generalizable AI-Generated Image Detection with DINOv3

2026-08-15

AI Summary

This study addresses the challenge of balancing data alignment and generalization in AI-generated image detection by proposing an Anchor-Regularized Adaptation method. This approach integrates low-rank adaptation to capture pixel-level artifacts while employing a frozen anchor classifier to preserve pretrained representational structures, thereby enabling complementary supervision across misaligned and aligned data. Implemented on the DINOv3 backbone, this mechanism effectively balances artifact cue capture with feature stability. Experimental results demonstrate that the proposed method achieves state-of-the-art performance across nine diverse benchmarks, significantly enhancing detection generalizability. Consequently, this work establishes a novel paradigm for reconciling detection accuracy with robustness in synthetic image forensics.

Abstract

Recent works in AI-generated image detection have shown that careful training data alignment can improve generalization by removing spurious correlations. However, linear probes on frozen DINOv3 representations achieve remarkably strong performance even when trained on misaligned datasets. Motivated by this result, we analyze the underlying rationale and the limits of this generalization. We find that frozen DINOv3 performs well because its decisions rely on features that faithfully represent the space of authentic images. At the same time, its final layer is less effective at capturing the subtle pixel-artifact cues that can be emphasized by aligned training data. We further observe that naively mixing aligned and misaligned data during adaptation improves sensitivity to such cues but at the cost of distorting the pre-trained representation, limiting generalization. To address this issue, we propose Anchor-Regularized Adaptation (ARA). We apply Low-Rank Adaptation to capture pixel-level artifacts while leveraging a frozen anchor classifier to avoid deviations from the original representation structure. This allows the model to exploit pixel-artifact cues without sacrificing generalization. Our method achieves state-of-the-art performance on nine diverse and challenging benchmarks, indicating that ARA enables complementary supervision from misaligned and aligned data for more effective detection.

Problem

• AI-generated image detection
• generalization
• representation distortion
• pixel-artifact cues
• data alignment

Innovation

• Anchor-Regularized Adaptation
• DINOv3
• Low-Rank Adaptation
• AI-Generated Image Detection
• Generalization