Avatar Kinect's Silent Scream Hints At Enormous Potential

Illustration for article titled Avatar Kinects Silent Scream Hints At Enormous Potential

The silent man contorts his face in expressions of agony and humor. A fine web of digital lines traces every movement. He grimaces, he smiles, his eyes go wide, his mouth gapes open in a silent scream.


Nearby a second video shows a strikingly similar cartoon adaptation of the silent screamer going through the same facial contortions.

This is the technology behind Avatar Kinect in action, an upcoming software update that will give the Xbox 360's motion-tracking camera array the ability to track not just the gross movements of the body, but the finer tics of smiles, of frowns, of raised eyebrows, of emotion.


For now the new technology will only be coming to one bit of software, the social chatting of Avatar Kinect. It will be used, Microsoft spokesman David Dennis tells Kotaku, to record the movements of a person's upper body and face and convert it into nearly real-time avatar movements and emotions.

"Kinect has some basic facial recognition already built into it for signing on," Dennis points out. "This takes that and blows that out and makes it more accurate. Instead of measuring a few points of the face. It maps the corner of your eyebrows. the corners of your mouth, the face."

In action, the demo appears to show the software mapping set points around the eyes, along the eyebrows, and around the mouth.

The newly-emotive avatars will be dropped into any of 15 digital environments along with up to seven friends to host chat sessions. The environments, Dennis says, includes tailgating, townhall meeting spaces and living rooms.


The idea is that people will be able to have real-time chats with up to seven friends, in an environment that can be more inventive, perhaps more comfortable, for people online.

The chats can even be recorded and uploaded to the Kinect Share website where they can be instantly shared on Facebook or downloaded to a computer for use on other sites, like YouTube. The current video length limit is ten minutes, Dennis said.


"It's meant to be a highly connected social experience," he said. "It could be a single person recording their daily musing or a group using it as a way to capture conversations or conduct a podcast."

While Microsoft doesn't yet have any announced plans for the facial-mapping technology to extend into more game-like experiences, it seems like a no-brainer.


Adding the ability to transmit real facial expressions in games like the upcoming avatar-based Full House Poker or even in triple-A titles, could add a new, interesting facet to multiplayer gaming.

Share This Story

Get our newsletter


It's the recording element that makes this a pretty neat idea.

I can see people making simple machinima with this, or more likely podcasts and YouTube shows (as suggested by Dennis).

Sure, it probably skews more towards the younger crowd, but the potential ease of creating this kind of content is still pretty neat.


Just to add onto this, could you imagine if somebody made a virtual film studio game with this? Take everybody who's been creating machinima with Halo or Gary's Mod, give them more realistic characters than the Xbox avatars with the included real-time motion capture and easy upload service. This could be huge.

Sorry, I got excited.