Deepfake Detection Generalization with Diffusion Noise

2026-04-15

AI Summary

Existing deepfake detection methods exhibit limited generalization when confronted with novel deepfake images generated by diffusion models. This work proposes an Attention-guided Noise Learning (ANL) framework that, for the first time, leverages the noise prediction from the denoising process of a pre-trained diffusion model as a supervisory signal. By employing an attention mechanism, ANL directs the detector to focus on global image inconsistencies, effectively capturing subtle forgery traces. The approach utilizes a frozen diffusion model as a regularizer, significantly enhancing cross-model generalization without incurring additional inference overhead. Experimental results demonstrate that ANL achieves state-of-the-art performance across multiple benchmarks, substantially improving detection accuracy—measured by ACC and AP—on images synthesized by unseen diffusion models.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Machine Learning: Deep Generative Models & Autoencoders
• Natural Language Processing: Fact-Checking / Misinformation Detection (NLP Focus)

Application Category

• Web Mining and Content Analysis: Large pretrained models with web data
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

Deepfake detectors face growing challenges in generalization as new image synthesis techniques emerge. In particular, deepfakes generated by diffusion models are highly photorealistic and often evade detectors trained on GAN-based forgeries. This paper addresses the generalization problem in deepfake detection by leveraging diffusion noise characteristics. We propose an Attention-guided Noise Learning (ANL) framework that integrates a pre-trained diffusion model into the deepfake detection pipeline to guide the learning of more robust features. Specifically, our method uses the diffusion model's denoising process to expose subtle artifacts: the detector is trained to predict the noise contained in an input image at a given diffusion step, forcing it to capture discrepancies between real and synthetic images, while an attention-guided mechanism derived from the predicted noise is introduced to encourage the model to focus on globally distributed discrepancies rather than local patterns. By harnessing the frozen diffusion model's learned distribution of natural images, the ANL method acts as a form of regularization, improving the detector's generalization to unseen forgery types. Extensive experiments demonstrate that ANL significantly outperforms existing methods on multiple benchmarks, achieving state-of-the-art accuracy in detecting diffusion-generated deepfakes. Notably, the proposed framework boosts generalization performance (e.g., improving ACC/AP by a substantial margin on unseen models) without introducing additional overhead during inference. Our results highlight that diffusion noise provides a powerful signal for generalizable deepfake detection.

Problem

• deepfake detection
• generalization
• diffusion models
• image synthesis
• photorealistic forgeries

Innovation

• diffusion noise
• deepfake detection
• generalization
• attention-guided learning
• pre-trained diffusion model