DiffusionPrint: Learning Generative Fingerprints for Diffusion-Based Inpainting Localization

2026-04-14

AI Summary

This work addresses the challenge that existing image forgery localization methods struggle to detect forgeries produced by diffusion-based inpainting, as such techniques reconstruct the entire image and thereby disrupt conventional forensic cues like camera sensor noise patterns. To overcome this limitation, the authors propose a patch-level contrastive learning framework that leverages the “generation fingerprint”—a shared artifact inherent to regions synthesized by diffusion models—as a self-supervised signal. By integrating MoCo-style contrastive learning with cross-category hard negative mining, the method trains a convolutional backbone to extract forensic features robust to latent-space decoding distortions. Additionally, a generator-aware classification head is introduced to enhance generalization. The approach significantly outperforms current state-of-the-art methods across diverse generative models, achieving up to a 28% improvement in localization accuracy on unseen mask types and demonstrating strong generalization to unknown generation architectures.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Machine Learning: Deep Generative Models & Autoencoders
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Web Mining and Content Analysis: Large pretrained models with web data
• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems

Abstract

Modern diffusion-based inpainting models pose significant challenges for image forgery localization (IFL), as their full regeneration pipelines reconstruct the entire image via a latent decoder, disrupting the camera-level noise patterns that existing forensic methods rely on. We propose DiffusionPrint, a patch-level contrastive learning framework that learns a forensic signal robust to the spectral distortions introduced by latent decoding. It exploits the fact that inpainted regions generated by the same model share a consistent generative fingerprint, using this as a self-supervisory signal. DiffusionPrint trains a convolutional backbone via a MoCo-style objective with cross-category hard negative mining and a generator-aware classification head, producing a forensic feature map that serves as a highly discriminative secondary modality in fusion-based IFL frameworks. Integrated into TruFor, MMFusion, and a lightweight fusion baseline, DiffusionPrint consistently improves localization across multiple generative models, with gains of up to +28% on mask types unseen during fine-tuning and confirmed generalization to unseen generative architectures. Code is available at https://github.com/mever-team/diffusionprint

Problem

• image forgery localization
• diffusion-based inpainting
• generative fingerprint
• forensic signal
• latent decoding

Innovation

• Diffusion-based inpainting
• generative fingerprint
• contrastive learning
• image forgery localization
• self-supervised forensics