Starting in 2020, the use of video conferencing services such as Google Meet and Zoom has increased. Due to the potential for improving the experience, Nvidia announces the solution in the form of the Maxine developer platform. Maxine with Nvidia GPU in the cloud offers a lot of interesting things.

One is the placement of faces to rotate the faces so that people look like they are facing each other. It aims to solve the very basic problems that most of us face. It’s difficult to see the other person because you don’t look at the camera while you’re talking.

But with Maxine, Nvidia says that not only can it look the same (I’ve seen Apple do it in FaceTime), but it can also rotate its face. This is strangely cool. Check out the video above to see the actual technology.

Feature of Maxine Video Conferencing

Maxine also offers AI video compression, which claims to use one-tenth the bandwidth of the H.254 video compression standard. Instead of broadcasting the entire screen over the Internet, Maxine analyzes a person’s key facial points and uses these facial points to “intelligently retouch” the face in the video to the other side. Do this with.

Another interesting feature that Maxine brings to video conferencing is Ultra HD, which allows you to convert low-definition video into HD video in real time. The video above shows how Nvidia works when converting a 360p feed to 720p. Maxine also improves performance in low light and can also automatically position the camera on the user’s face.

Other features offered by Maxine include language translation, default backgrounds like zoom in, and virtual avatars like Animoji

Given the growing popularity of video conferencing services around the world, we welcome the improvements Nvidia Maxine claims to bring. However, in order to actually benefit from any of these technologies, it is imperative that the major players partner with Nvidia to incorporate Maxine.