从外表和行为来检测假冒视频（multimedia）---用户6869393

合成生成的音频和视频——所谓的“深度伪造”——继续吸引着计算机图形和计算机视觉社区的想象力。与此同时，技术的民主化仍然令人担忧，因为它有可能破坏民主选举，造成小规模的大规模舞弊，助长虚假信息宣传，并制造非经双方同意的色情内容。我们描述了一种基于生物识别技术的法医技术，用于检测面部置换深度假货。该技术结合了基于面部识别的静态生物特征和基于面部表情和头部运动的行为生物特征，其中行为嵌入是使用带有度量学习目标函数的CNN来学习的。我们通过几个大规模的视频数据集，以及在野外的深度假数据集，展示了这种方法的有效性。

原文题目：Detecting Deep-Fake Videos from Appearance and Behavior

原文：Synthetically-generated audios and videos – so-called deep fakes – continue to capture the imagination of the computer-graphics and computer-vision communities. At the same time, the democratization of access to technology that can create sophisticated manipulated video of anybody saying anything continues to be of concern because of its power to disrupt democratic elections, commit small to large-scale fraud, fuel dis-information campaigns, and create non-consensual pornography. We describe a biometric-based forensic technique for detecting face-swap deep fakes. This technique combines a static biometric based on facial recognition with a temporal, behavioral biometric based on facial expressions and head movements, where the behavioral embedding is learned using a CNN with a metric-learning objective function. We show the efficacy of this approach across several large-scale video datasets, as well as in-the-wild deep fakes.

原文作者：Shruti Agarwal (1), Tarek El-Gaaly (2), Hany Farid (1), Ser-Nam Lim (2) ((1) Univeristy of California, Berkeley, Berkeley, CA, USA, (2) Facebook Research, New York, NY, USA)

原文链接：https://arxiv.org/abs/2004.14491