The program can be produce hyper-realistic AI-generated animations with just a single photo reference and an audio clip, prompting fears that Microsoft has unwittingly released a deepfaking program upon the broad masses.

According to blog post on Microsoft’s website, the company says:

“We introduce VASA, a framework for generating lifelike talking faces of virtual characters with appealing visual affective skills (VAS), given a single static image and a speech audio clip. Our premiere model, VASA-1, is capable of not only producing lip movements that are exquisitely synchronized with the audio, but also capturing a large spectrum of facial nuances and natural head motions that contribute to the perception of authenticity and liveliness,” Microsoft promotes.

“The core innovations include a holistic facial dynamics and head movement generation model that works in a face latent space, and the development of such an expressive and disentangled face latent space using videos. Through extensive experiments including evaluation on a set of new metrics, we show that our method significantly outperforms previous methods along various dimensions comprehensively.

“Our method not only delivers high video quality with realistic facial and head dynamics but also supports the online generation of 512×512 videos at up to 40 FPS with negligible starting latency. It paves the way for real-time engagements with lifelike avatars that emulate human conversational behaviors,” they explain.

Our method is capable of not only producing precious lip-audio synchronization, but also generating a large spectrum of expressive facial nuances and natural head motions. It can handle arbitrary-length audio and stably output seamless talking face videos. 

Microsoft added

The tech giant then provides a myriad of examples and applications as to what can be created, also showing that it effectively works with sketches and art pieces, for example.