arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2007.10466eess.IVcs.CV
收藏

GAN生成图像的检测、属性和定位

Detection, Attribution and Localization of GAN Generated Images

University of California, Santa Barbara(加州大学圣塔芭芭拉分校);Mayachitra, Inc.(Mayachitra公司);University of California(加州大学)

Michael Goebel, Lakshmanan Nataraj, Tejaswi Nanjundaswamy, Tajuddin Manhar Mohammed, Shivkumar Chandrasekaran, B. S. Manjunath

摘要:

生成对抗网络(GANs)的最新进展导致创建了逼真的数字图像,这对人类或计算机的检测构成了重大挑战。GANs用于各种任务,从修改图像的小属性(StarGAN [14]),在图像对之间传输属性(CycleGAN [91]),以及生成全新的图像(ProGAN [36],StyleGAN [37],SPADE/GauGAN [64])。在本文中,我们提出了一种新的方法来检测,属性和定位GAN生成的图像,结合图像特征与深度学习方法。对于每幅图像,在不同方向(水平、垂直和对角线)的RGB通道的邻域像素上计算共生矩阵。然后在这些特征上训练深度学习网络,以检测、归属和定位这些GAN生成/操纵的图像。在5个GAN数据集上对我们的方法进行了大规模评估,包括超过276万张图像(ProGAN,StarGAN,CycleGAN,StyleGAN和SPADE/GauGAN),在检测GAN生成的图像方面显示了有希望的结果。

英文摘要:

Recent advances in Generative Adversarial Networks (GANs) have led to the creation of realistic-looking digital images that pose a major challenge to their detection by humans or computers. GANs are used in a wide range of tasks, from modifying small attributes of an image (StarGAN [14]), transferring attributes between image pairs (CycleGAN [91]), as well as generating entirely new images (ProGAN [36], StyleGAN [37], SPADE/GauGAN [64]). In this paper, we propose a novel approach to detect, attribute and localize GAN generated images that combines image features with deep learning methods. For every image, co-occurrence matrices are computed on neighborhood pixels of RGB channels in different directions (horizontal, vertical and diagonal). A deep learning network is then trained on these features to detect, attribute and localize these GAN generated/manipulated images. A large scale evaluation of our approach on 5 GAN datasets comprising over 2.76 million images (ProGAN, StarGAN, CycleGAN, StyleGAN and SPADE/GauGAN) shows promising results in detecting GAN generated images.

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。

↑