Skip to main content
Lasque Tiarc

Microsoft’s new AI makes Mona Lisa rap. How does it work?

Vaseline 2 months ago

(CNN) — Microsoft unveils new AI technology that brings the Mona Lisa to life

Last week, Microsoft researchers introduced a groundbreaking artificial intelligence model called VASA-1, which can animate a still image of a face with an audio clip to create a realistic video of the person speaking. This technology enables convincing lip synchronization and natural facial and head movements.

In a captivating demonstration, researchers showed how they animated the iconic Mona Lisa to recite a hilarious rap from actress Anne Hathaway. The results of this AI model are both entertaining and surprisingly realistic, expanding the possibilities for education, accessibility, and even virtual companions for individuals. However, concerns arise about possible abuse and deception as this technology can be used to impersonate real people.

As the tech industry continues to develop tools for generating AI images and videos, experts warn of the risks of misinformation and potential disruptions to the creative industries. While Microsoft has no immediate plans to release the VASA-1 model to the public, they emphasize the importance of responsible use and regulatory compliance.

Microsoft’s innovative AI model is trained on a variety of videos showing people’s faces talking, allowing it to capture natural facial movements such as lip sync, expressions, glances and blinks. While there are still indications that the videos are machine-generated, Microsoft believes their model excels at creating realistic avatars that mimic human conversational behavior in real time.

The unveiling of this technology opens exciting possibilities for the future, but also underlines the need for careful consideration and ethical use of AI developments.