In the early days of the Web, internet cafes used to be very popular. People would go to have a coffee and surf the web together. It was all very exotic.
Of course it soon became not exotic at all, as everyone got browsers on their PCs at home. As a social practice, browsing the web because less like going to the movies and more like watching TV.
Now we are about to go through a similar set of transitions for “physically” hanging out with people in shared virtual reality. I’m not talking about the flavor of VR where you sit down in a chair and just pretend to walk around. I mean the more interesting kind, where you physically walk around with your own body, wearing a headset, but with no trailing wires to encumber you.
At first, doing this in high quality is going to be somewhat expensive, and therefore exotic. So I suspect you will see the equivalent of internet cafes popping up in your town.
Then after a few years, when the prices go down and walking around in VR just becomes a widespread capability found in the home, those “VR Cafes” will become a thing of the past, something to look back on with nostalgia.
Gosh, I almost miss them already. 🙂
Maybe you would know– has anyone ever worked on a VR system that uses a Kinect to see what objects are around you, and then replaces them with something from the game setting? I’m imagining a game set in the future, that replaces all cars with landspeeders, and all people with aliens, for example. It’s VR, not augmented reality– everything you see is generated by the computer– but you won’t accidentally walk into things in the real world, and you can touch anything you can see.
Various people have been working on different pieces of that. For example, people at Microsoft have done a great job at constructing a 3D ‘sculpture’ of a scene over time from successive depth images captured with a depth camera.
The harder part is reliably interpreting that 3D virtual object from its shape, so that you know whether you are looking at cars, people, etc. The recognition problem is easiest when things don’t move (like a couch), more difficult when they move rigidly (like a ball), and hardest when they move non-rigidly (as is the case for people, dogs, etc).
If you just want to reliably place virtual walls, tables, etc., where the real walls, tables, etc. are, then it’s much easier, since you don’t need to know what those things are — you can just accumulate a 3D scene over time using your depth camera (or using a set of depth cameras all working together).