Face camouflage method and system based on expression action migration

The invention provides a face camouflage method and system based on expression action migration, and the method comprises the steps: extracting a driving video which is made by a third party according to a random instruction transmitted by a face recognition system, obtaining a difference value between a key point of each frame of image in the driving video and a key point of a target face, and generating an action video of the target face, therefore, the action of the third party in the driving video is migrated to the target face, namely, the appearance of the target face is reserved, the motion characteristics of the face of the third party in the driving video are added, an action video of the target face is obtained, and face recognition is completed. 本发明提供了一种基于表情动作迁移的人脸伪装方法及系统，通过提取第三者根据人脸识别系统发出的随机指令所做的驱动视频，将驱动视频中每一帧图像的关键点与目标人脸的关键点取差值，生成目标人脸的动作视频，从而将第三者在驱动视频中的动作迁移到目标人脸中，即保留了目标人脸的外观，加上驱动视频中第三者人脸的运动特性，得到目标人脸的动作视频，完成人脸识别。