arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2105.04580cs.CVcs.LG
收藏

面向开放世界GAN生成图像的发现和属性

Towards Discovery and Attribution of Open-world GAN Generated Images

University of Maryland, College Park(马里兰大学帕克分校)

Sharath Girish, Saksham Suri, Saketh Rambhatla, Abhinav Shrivastava

摘要:

随着生成对抗网络(GANs)的最新进展,媒体和视觉取证必须开发能够识别图像并将其归因于生成它们的模型的检测器。现有的工作表明,属性的图像,以其相应的GAN源具有高精度。然而,这些工作仅限于闭集场景,无法推广到在训练时间内看不见的GANs,因此无法随着新GANs的稳定涌入而扩展。我们提出了一种迭代算法,通过利用所有GANs在其生成的图像上留下不同指纹的事实,来发现从以前看不见的GANs生成的图像。我们的算法包括多个组件,包括网络训练,分布外检测,聚类,合并和细化步骤。通过大量的实验,我们表明,我们的算法发现看不见的GANs具有很高的准确性,也推广到GANs训练看不见的真实数据集。我们还将我们的算法应用于在线方式的GANs的归属和发现,以及更标准的真/假检测任务。我们的实验证明了我们的方法在发现新的GANs方面的有效性,并且可以在开放世界的设置中使用。

英文摘要:

With the recent progress in Generative Adversarial Networks (GANs), it is imperative for media and visual forensics to develop detectors which can identify and attribute images to the model generating them. Existing works have shown to attribute images to their corresponding GAN sources with high accuracy. However, these works are limited to a closed set scenario, failing to generalize to GANs unseen during train time and are therefore, not scalable with a steady influx of new GANs. We present an iterative algorithm for discovering images generated from previously unseen GANs by exploiting the fact that all GANs leave distinct fingerprints on their generated images. Our algorithm consists of multiple components including network training, out-of-distribution detection, clustering, merge and refine steps. Through extensive experiments, we show that our algorithm discovers unseen GANs with high accuracy and also generalizes to GANs trained on unseen real datasets. We additionally apply our algorithm to attribution and discovery of GANs in an online fashion as well as to the more standard task of real/fake detection. Our experiments demonstrate the effectiveness of our approach to discover new GANs and can be used in an open-world setup.

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。

↑