VASA-1, Microsoft’s new AI model, is finally arriving. Here’s what he can do, it seems like science fiction.
Since the beginning, Microsoft wanted to establish itself as one of the companies that most want to aim for on artificial intelligence. And indeed this was the case, thanks to its systems that give users the opportunity to access a myriad of benefits quickly and easily. So you can enjoy the latest generation technologies that you can test and use for free both for leisure and for work.
And obviously it’s not over yet, on the contrary. As emerged during the presentation that just took place, in fact, Microsoft now wants to go further with its new AI system. That is VASA-1a fantastic tool that will give you the opportunity to exploit even greater potential right away. The images look like science fictionand how realistic they are, after testing everything you will be speechless. And for the less confident, even a little scared of the potential of this tool.
Microsoft launches VASA-1: here’s what it is and what it will be used for
VASA-1 is the new artificial intelligence system presented by Microsoft which immediately promises to be revolutionary. Because of its advanced technical specifications which will give users the opportunity to enjoy a tool that has never been seen before. And it could be very useful.
The idea is to be able to generate realistic talking faces of virtual or real characters. Using a single static image and a short audio clip. And the craziest thing is that this model has the ability to realistically generate output even if what a user wants was not in the model’s training deployment.
Think for example about the fact that, according to Microsoft, it will also be able to manage artistic photos, audio tracks with songs, speeches not in English and much more. So here it is, if you want for example make the Mona Lisa speak, you can do it! VASA-1 will generate 512×512 video frames at 45 fps in offline batch processing mode, while streaming drops to 40 fps with 170 ms latency.
We’ll see how this bot will be received by users. To date it is not yet available, as the company has stated that there are no plans to release online demos or APIs. At least until we are certain that the technology will come used responsibly and in compliance with the most appropriate regulations.
