Hacker News Reader

Comment by burgerbrain | Hacker News Reader

"Minority Report" interfaces look really cool when demoed, but I have yet to see a solution to gorilla arm syndrome. No matter how good the technology gets, nobody is going to want to wave their arms around in front of them for anything even close to "an extended amount of time".

This technology has a lot of potential, but this particular application is a dead end.

View on HN

Another commenter called out a use I hadn't thought of: operating media center PCs without a mouse or remote. Since you only need brief interactions, the fatigue wouldn't a big deal.

Of course, in order not to change the channel, you'll have to stand very still...

I'm not sure if you are joking here or not -- but, if you've played with the Kinect on a 360, you'd note that it can recognize your hands, and only when you hold your hands out for a set period of time (2-3 seconds, cannot recall exactly) the Kinect wakes up and begins to allow you to input controls.

I was joking. I failed to locate the correct reference to the Hitchhikers Guide to The Galaxy, but I am sure there is a joke on gesture recognition somewhere in there.

Personally, I'd rather tell my media centre PC to change channel with a voice command than wave my hand at it. Seems more natural to me. I'm sure flicking through channels would be easy enough with some kind of motion commands, but navigating to specific channels, how would you do that? I guess you could have a specific gesture for each individual channel, but even then, it would me much easier to just say the name of a chennel. Another idea would be a specific gesture to go to the list of channels, which you could then scroll through and select, but this is still more work then voice commands. I'm not saying this technology sucks, I think it's fantastic, but it doesn't solve every UI problem.

Ive read a commenter here in HN remark on how the Voice recognition of the Kinect is something that is seriously being slept on/not paid enough attention too.

That's because it's just 4 microphones. You STILL have to do the speech recognition yourself, and that's tough.

The new part of the Kinect is the 3D stuff, so that's what people are playing with.

That's odd, the XBox SDK had a speech recognition library build in to it. I presumed that shipped as part of the XBox 360 SDK?

This is only a problem when we use it for a 2d display. Use it with AR glasses and voila, you have freedom.

No need to wave in a direct direction to do unrelated motions, when you can do what you want in front of you, the "usual" way.

It's only a dead end if you're assuming a situation where people would be doing this for long periods of time. If you're just switching viewports, changing the view of a CAD drawing or performing other infrequent tasks then this sort of user interface would be ok, and quite intuitive.

I haven't tried it but I think I'd appriciate the exercise. Anything is better than sitting a desk staring at a screen. I wish computing could involve walking around a room too.

But you can use it from a desk! Imagine sitting at a smallish desk with a keyboard and a projector/large screen a few feet away. Gestures whenever you want them (if you wrote the software right, you could even do most of them with your elbows on the desk), typing whenever you need, voice commands the rest.

I don't know what the near point on the kinect is, but if they could drop it down ~foot, then you could use it at your desk right now. It'll be nice for some image/CAD stuff. Shame it's not really sensitive enough to pick out wrist/finger movement. Then you could use your entire desk as your touchpad! Magic-pad eat my dust.

Reply on news.ycombinator.com