Introducing VASA-1, the groundbreaking AI tool from Microsoft Research Asia that can transform a still image or drawing of a person along with an audio file into a realistic talking face in real time. This innovative tool can create facial expressions, head movements, and lip sync to match speech or songs. The results are so convincing that they can easily deceive viewers into believing they are real.
Although some may notice slight robotic movements in the examples, it’s evident that this technology could potentially be used to create deepfake videos of individuals. The researchers are cautious of this misuse and have refrained from releasing any online demos or related products until they can ensure…
2024-04-20 02:00:52
Originally published on www.engadget.com