[论文解读] FIND: A Simple yet Effective Baseline for Diffusion-Generated Image Detection

Jie Li, Yingying Feng+4 more|arXiv (Cornell University)|Mar 15, 2026

Advanced Neuroimaging Techniques and Applications被引用 0

一句话总结

FIND 通过给真实图像添加高斯噪声并将带噪版本标记为合成数据来训练二分类器，消除了对重建模型的需求，在 GenImage 上达到最先进的准确性和速度。

ABSTRACT

The remarkable realism of images generated by diffusion models poses critical detection challenges. Current methods utilize reconstruction error as a discriminative feature, exploiting the observation that real images exhibit higher reconstruction errors when processed through diffusion models. However, these approaches require costly reconstruction computations and depend on specific diffusion models, making their performance highly model-dependent. We identify a fundamental difference: real images are more difficult to fit with Gaussian distributions compared to synthetic ones. In this paper, we propose Forgery Identification via Noise Disturbance (FIND), a novel method that requires only a simple binary classifier. It eliminates reconstruction by directly targeting the core distributional difference between real and synthetic images. Our key operation is to add Gaussian noise to real images during training and label these noisy versions as synthetic. This step allows the classifier to focus on the statistical patterns that distinguish real from synthetic images. We theoretically prove that the noise-augmented real images resemble diffusion-generated images in their ease of Gaussian fitting. Furthermore, simply by adding noise, they still retain visual similarity to the original images, highlighting the most discriminative distribution-related features. The proposed FIND improves performance by 11.7% on the GenImage benchmark while running 126x faster than existing methods. By removing the need for auxiliary diffusion models and reconstruction, it offers a practical, efficient, and generalizable way to detect diffusion-generated content.

研究动机与目标

• 动机： Robust 检测扩散生成的图像，解决基于重建的检测器在泛化与效率方面的空缺。
• 提出一种无重建基线，通过对真实与合成图像之间的分布差异进行噪声增强来实现。
• 从高斯拟合角度理论证明为何带噪的真实图像在分布上与扩散生成数据相似。
• 在 GenImage 上通过多种生成器实证展示对比基线的高准确性与极高速度优势。

提出的方法

• 在扩散重建的上下文中假设真实图像在高斯分布拟合中比合成图像更难拟合。
• 在训练时对真实图像加入高斯噪声，并将这些带噪样本标记为合成以诱导一个以分布为中心的判别特征。
• 在仅使用真实、带噪真实和合成图像的情况下训练一个简单的二分类器，不使用任何重建或辅助扩散模型。
• 证明噪声增强会降低 Fisher 信息量，使真实-噪声分布在分类器层面更接近于合成分布。
• 在 GenImage 上使用 ViT-B/16 与 RN50 主干进行 FIND 的评估，并与基于重建的基线进行对比。
• 证明推断不需要重建，且速度显著快于现有方法。

Figure 1: Comparison between FIND and previous noise-based methods. (a) DIRE utilizes complete reconstruction by adding noise and then denoising it with a Diffusion Model (DM) in multiple steps; (b) LaRE 2 compresses the noise addition and denoising steps into a single step within the latent space;

实验结果

研究问题

• RQ1 一个无重建的二分类器是否能有效将真实图像与扩散生成图像区分开？
• RQ2 对真实图像进行高斯噪声增强是否会使其分布向扩散生成数据靠拢，从而提升检测效果？
• RQ3 相较于基于重建的检测器，FIND 在多种扩散生成器上的泛化性如何？

主要发现

• FIND 在 GenImage 的 RN50 与 ViT-B/16 主干上均实现高于先前方法的平均准确率。
• 使用 ViT-B/16 时，FIND 在各生成器上的平均准确率为 88.35%，超越 FatFormer 与 LaRE 2。
• FIND 取消对重建模型和基于扩散的重建需求，相比最新方法推断速度提升约 126 倍。
• 对真实图像进行随机高斯噪声训练显著提升性能，尤其是将真实图像噪声与真实和合成数据的噪声结合时。
• 最优噪声幅度（epsilon）为 50，达到图像保真与判别能力的平衡。
• FIND 显示出稳定的跨生成器表现，在 GenImage 的八个扩散生成器上具有良好的泛化性。

Figure 2: The training framework of FIND. Gaussian noise is added to real images, and these perturbed versions are labeled as synthetic within the training process. This enables FIND to mitigate the reconstruction model dependency, and learn the core distributional differences between real and synth