AI has reached the point where it can deceive us. With a few hundred hours of reference footage, neural networks can produce a video of you saying words you never spoke, things you never did. Even your closest friends could be convinced by the fidelity of these fabricated lies. These videos are known as deep fakes.
This visualization was a small segment of a large multimedia story on deep fakes told by my colleagues at CNN here.
For the facial analysis I designed some data elements for the eyes, nose and mouth to convey that they were being scanned to create the deep fakes.
This segment was designed to illustrate how the computer analyzes a group of photos so it can then map the corresponding face to the one that’s already there.