Facial synthesis in augmented reality content for advertisements

A system comprising:

• a processor;
• and
• a memory including instructions that, when executed by the processor, cause the processor to perform operations comprising:
• generating, by one or more hardware processors, sets of source pose parameters, the sets of the source pose parameters comprise positions of representations of a head of a source actor and facial expressions of the source actor in frames of a source media content, wherein generating the sets of source pose parameters is performed using a shared encoder network, the shared encoder network comprises a first deep convolutional neural network;
• receiving, by the one or more hardware processors, at least one target image, the at least one target image including representations of a target head and a target face of a target entity;
• generating based at least in part on the sets of source pose parameters, by the one or more hardware processors, an output media content, each frame of the output media content including an image of the target face in at least one frame of the output media content, the image of the target face being modified based on at least one of the sets of the source pose parameters to mimic at least one of the positions of the head of the source actor and at least one of the facial expressions of the source actor, wherein generating the output media content is performed using a first decoder network, the first decoder network comprises a second deep convolutional neural network;
• and
• causing playback of an online advertisement based on a duration of the online advertisement in a messaging client application on a client device, wherein the online advertisement comprises a video, the video not being able to be skipped or paused during playback of the video, the online advertisement further comprises an image, the image not being able to be hidden during the duration of the online advertisement, the image being selected based on respective properties comprising a name, brand name, or headline.

A non-transitory computer-readable medium comprising instructions, which when executed by a computing device, cause the computing device to perform operations comprising:

• generating, by one or more hardware processors, sets of source pose parameters, the sets of the source pose parameters comprise positions of representations of a head of a source actor and facial expressions of the source actor in frames of a source media content, wherein generating the sets of source pose parameters is performed using a shared encoder network, the shared encoder network comprises a first deep convolutional neural network;
• receiving, by the one or more hardware processors, at least one target image, the at least one target image including representations of a target head and a target face of a target entity;
• generating based at least in part on the sets of source pose parameters, by the one or more hardware processors, an output media content, each frame of the output media content including an image of the target face in at least one frame of the output media content, the image of the target face being modified based on at least one of the sets of the source pose parameters to mimic at least one of the positions of the head of the source actor and at least one of the facial expressions of the source actor, wherein generating the output media content is performed using a first decoder network, the first decoder network comprises a second deep convolutional neural network;
• and
• causing playback of an online advertisement based on a duration of the online advertisement in a messaging client application on a client device, wherein the online advertisement comprises a video, the video not being able to be skipped or paused during playback of the video, the online advertisement further comprises an image, the image not being able to be hidden during the duration of the online advertisement, the image being selected based on respective properties comprising a name, brand name, or headline.