Microsoft’s VASA-1 can deepfake a person with one photo and one audio track
YouTube videos of 6K celebrities helped train AI model to animate photos in real time.
https://arstechnica.com/?p=2018178
Discover the world at Altruu, The Discovery Engine