Mark Zuckerberg, the CEO of Meta, has reignited interest in the metaverse in a virtual reality interview that has left many impressed. During a conversation with Russian-American computer scientist and podcaster Lex Fridman, Zuckerberg presented super-realistic avatars in what can be considered the first live demo of Meta's 'Pixel Codec Avatars' project. This project aims to allow virtual reality headsets, such as the Quest Pro glasses, to analyze people's faces and create real-time avatars that replicate their facial gestures while they speak.
The interview, which took place entirely in virtual reality, allowed Zuckerberg and Fridman to come "face to face" despite being thousands of miles apart. "We are hundreds of miles away from each other in physical space, but it feels like we are in the same room because we appear as photorealistic 3D Kodak avatars with spatial audio," the podcaster commented at the beginning of the interview.
"This technology is incredible and I believe it is the future of how humans connect with each other in a deeply meaningful way on the Internet. These avatars can capture many of the nuances of facial expressions that humans use to communicate and move with each other "Friedman commented.
Creating these highly detailed avatars requires a lot of resources and time these days. The CEO of Meta explained that, to carry out this interview, both protagonists made some scans for “this research project we have at Meta called Kodak Avatars.” The system, once the faces have been scanned in addition to many different expressions, builds each of the faces and bodies that are then collapsed into a Kodak that then, “when the headphones are on your head, sees your face, sees your expression and It can basically send a scrambled version of what you're supposed to watch.”
Zuckerberg also highlighted the efficiency of this technology in terms of bandwidth to transmit high-quality 3D content. Although it was not specified whether the avatars were generated on the Quest Pro glasses or on an external computer, the fluidity of the conversation in virtual reality is impressive.
For now, the technology has limitations, since faces and facial expressions must be scanned in a specific system. To simplify it and make it accessible to citizens, Zuckerberg explained that they have a project they are already working on: scanning through mobile phones. “You just have to pick up the phone, wave it in your face for a couple of minutes, say a few sentences and make a bunch of facial expressions.” This technology is expected to be commercially available in about three years.
In the interview, Zuckerberg pointed out that another version of the improvements that will be transferred to the metaverse is how it will intersect with artificial intelligence (AI). “At some point, I think you can create an AI version of yourself that can interact with people not after you die, but while you are here to help people fulfill this desire to interact with you and your desire to build a community,” he commented.
To make it. clarifies the CEO of Meta, there are still “many things we need to do right: we are not yet ready to launch the version in which a creator can create a version of himself, but we are starting to experiment with it in terms of launching a series of AI that people can interact with in different ways,” he added.
The CEO of Meta resumed the conversation about the metaverse after the Meta Connect 2023 conference, in which the Meta Quest 3 and other news related to artificial intelligence were presented.
This presentation of realistic avatars seeks to calm the criticism that Zuckerberg's first avatar provoked in the metaverse, a couple of years ago. This advancement in avatar imagery shows Meta's ambition in the realm of virtual reality and the metaverse, which goes beyond immersive video calls.
