Who here every time joins a teleconference or video call turns off the camera so the caller can't see your face? Why did you do it, was it because you were shy or not ready to wake up with a new face? Coming soon, you can make video calls with your “own face” and voice without using a webcam. How?
Of course with a generative artificial intelligence system. Recently, Microsoft Research has shown VASA, which is an AI system that can generate hyper-realistic human faces along with emotions on the face. Not only that, VASA can also regenerate the human voice to speak like a human.
VASA-1, which is Microsoft's first vernacular model for AI like this, is said to be able to generate video with a resolution of 512×512 with a frame rate (FPS) of 40fps. This study states that the use of VASA as a character for teleconferences will be temporary but offline generation is just a test of generating characters with voices that have realistic speech performance and facial expressions.
Microsoft said that VASA is still not offered to the public because it is worried that there will be individuals who misuse this technology for criminal purposes. Although fraud detection systems have been developed, Microsoft is not yet ready to give the public access to VASA-1. Is this a generative AI technology you want to use? If it is limited to Microsoft Teams only and has to be paid at a high price, will you subscribe anyway?