Not a member yet? Register for full benefits!

3D Cameras will make Virtual Worlds Easier to Use?

At the Metaverse Roadmap meeting, Feb 15th, in Stanford, California, Mitch Kapor, the founder of Lotus 1-2-3, and primary investment backer in Second Life, announced plans for a new mainstream VR interface, which he is backing.

"I'm obsessed with what's going to make these things easier to use," Kapor said, his face lit with excitement. "I think a piece of hardware is involved."

His idea is not a new one; in fact it has been a staple of cyberpunk for quite some time.

The top surface of the computer is smooth except for a fisheye lens, a polished glass dome with a purplish optical coating. Whenever Hiro is using the machine, this lens emerges and clicks into place, its base flush with the surface of the computer. ? The lens can see half the universe ? the half that is above the computer, which includes most of Hiro. In this way, it can generally keep track of where Hiro is and what direction he?s looking in.

- Snow Crash, Page: 21

The idea Mitch Kapor suggested was to use a 3D camera array. Dissimilar to the 3D cameras used to provide QuickTime VR experiences, these would involve a bank of cameras focussed on one point in space - the space the user occupies.

These cameras form an input device, in conjunction with your body. They triangulate, to a degree not possible with a 2D device like eye-toy, the precise position of parts of your body as you move them, able to filter out background movement, which does not show up in all cameras.

Thus, to raise your avatar's hand, simply lift up your own. To nod your avatar's head, simply nod your own.

He stated that the today's graphical user interface standard input, a mouse and keyboard, don't really make that much sense, (no arguing with that) but that [most] people don't question it because there's no alternative.

But he pointed to the futuristic graphical display made famous in Steven Spielberg's Minority Report that allowed Tom Cruise's character to move information around on a screen with his hands.

"Look at Minority Report," Kapor said. "I'm here to say that [technology is] going to be real in the next [few] years."

Eye-Toy, one of the current, 2D camera systems

He predicted that 3D cameras, which would be built into computers much like regular 2D cameras are today, will be available in as little as 12 months.

Kapor said he wasn't clear on what the interface would be like, but he suggested it could be based on something like that of the Segway, in which users move their body forward and the Segway goes forward, backward to go backward, and so forth.

"So, if I look to the left in the real world, I just want my avatar to look to the left," Kapor said. "If I smile, I just want my avatar to smile. The cameras should be good enough to pick that up. I think we're going to see an amazing jump in the sense of presence."

Another experiential improvement the cameras could offer, he said, could be a better way to edit 3D objects.

"It's going to change how editing is done in 3D worlds," he said, "if you can reach in and grab the handles of an object and pull them out with your hands and extend or change the shape of the object."

To demonstrate the value of the cameras, Kapor said he has a prototype already and is planning, within a few months, to start putting up a series of videos shot with it onto YouTube.

Known Drawbacks


No interface is perfect; each has its own strengths and weaknesses. Each interface has a section of the userbase that can utilise it, and a segment of the userbase which can not.

In the case of a 3D bodily position system, the segment of the userbase which is going to have the most trouble working with it, consists of those with motor-neuron related or muscle wasting conditions.

With such physical manifestations, the virtual environment is often a blessed refuge from the physical (also true for just about every major disabling issue).

So, taking for an example, motorneuron disease (MND). MND isn't a single condition, but is an umbrella term for a collection of progressive neurological disorders that destroy motor neurons, making it a good example to take.

Motor neurons are the cells that control essential voluntary muscle activity such as speaking, walking, breathing, and swallowing.

When there are disruptions in these signals, the muscles do not work properly; one result can be uncontrollable twitching known as fasciculations. Another can be a stiffness of the limbs which leads to overreaction - jerkiness of movement, clumsiness.

The more astute reader can see where this is going: Up until now, virtual environments have been places where it is possible to escape the social implications of such movements.

If there are 3D cameras picking up every little movement of the upper body, and transferring it to the avatar, this degree of escape is no-longer possible.

Hectic Households

The systems are going to struggle in a hectic house as well. In fiction, the user always stays with them until they log out. In reality, what happens if the user has to use the lavatory, or answer the door? How does the system respond to that kind of swift, disappearing from sight movement? Is the avatar flung hard into a wall, trying to copy the sudden speed of dashing off?

Slightly more insidious. Picture this scenario: You are navigating in your favourite virtual space, gesturing to your friends, walking with them down a VR-based lane, when suddenly your partner leans across you. Suddenly, a second set of hands, attached to arms, attached to a second upper body is in sight of all the cameras. Two sets of bodies, controlling one avatar. The ensuing chaos, will be interesting from the point of everyone else present with your avatar.

Same problem from a different angle: Raise your hand if you have a child that loves to lean over and play with your keyboard whilst you're working, or a dog who repeatedly thumps the play ball on the desk, or cat who walks between you and the PC, regularly.

Ok, is there anyone left without a hand up? You can put them down now. Each and every one of these is going to confuse a camera-based system, as soon as they are close enough to be seen in all cameras, and so treated as a part of you?a part of you with no connection to your brain whatsoever, and usually with not much interest to the chaos they have caused with the virtual space.


So, will a 3D camera system make virtual worlds easier to use?

Well, the answer is, as it has to be, Yes, for a segment. That segment may well be the largest segment, at least of several potential markets, but its not the only segment.

It is a decisive step forwards if it can be pulled off well, but it is important that we never concentrate on just one possible interface method. Different physical circumstances should be kept in mind, lest we remove the greatest asset VR offers - a way to remove that which is otherwise beyond your control.


Mitch Kapor: 3D cameras will make virtual worlds easier to use

Staff Comments


Untitled Document .