arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2408.09153cs.CV
收藏

CLIP特征是否是通用合成图像溯源所需的全部?

Are CLIP features all you need for Universal Synthetic Image Origin Attribution?

University of Florence(佛罗伦萨大学);City, University of London(伦敦城市大学);Queen Mary, University of London(伦敦大学Queen Mary)

Dario Cioni, Christos Tzelepis, Lorenzo Seidenari, Ioannis Patras

AI总结:

针对合成图像滥用威胁,本文提出利用大型预训练基础模型特征的框架,在Open-Set场景下实现跨多种生成模型的合成图像溯源,在低数据量下性能超越现有方法且泛化性更好。

摘要:

Diffusion Models在视觉合成方面的持续改进催生了许多新的有趣的合成图像用例,但也引发了对其潜在滥用的担忧,这构成了重大的社会威胁。为解决此问题,需要检测伪造图像并将其归因于其源模型,且鉴于新生成器频繁发布,实际应用需考虑Open-Set场景,即某些模型在训练时未被见过。现有取证技术要么局限于Closed-Set设置,要么仅针对GAN生成的图像,依赖于脆弱的基于频率的“指纹”特征。相比之下,我们提出了一个简单而有效的框架,该框架融合了来自大型预训练基础模型的特征,以对各种生成模型(包括Diffusion Models)生成的合成图像进行Open-Set溯源。我们表明,即使在低数据量情况下,我们的方法也能实现卓越的溯源性能,超越了现有方法,并在从多种架构获取的图像上具有更好的泛化能力。我们在以下网址公开代码:https://github.com/ciodar/UniversalAttribution。

英文摘要:

The steady improvement of Diffusion Models for visual synthesis has given rise to many new and interesting use cases of synthetic images but also has raised concerns about their potential abuse, which poses significant societal threats. To address this, fake images need to be detected and attributed to their source model, and given the frequent release of new generators, realistic applications need to consider an Open-Set scenario where some models are unseen at training time. Existing forensic techniques are either limited to Closed-Set settings or to GAN-generated images, relying on fragile frequency-based "fingerprint" features. By contrast, we propose a simple yet effective framework that incorporates features from large pre-trained foundation models to perform Open-Set origin attribution of synthetic images produced by various generative models, including Diffusion Models. We show that our method leads to remarkable attribution performance, even in the low-data regime, exceeding the performance of existing methods and generalizes better on images obtained from a diverse set of architectures. We make the code publicly available at: https://github.com/ciodar/UniversalAttribution.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑