Slashdot Mirror
Wearable Motion Capture
AnonymousHack writes "Swiss and MIT researchers have developed a wearable kit that will capture your every move for mapping onto a virtual character. It's almost as accurate as the camera-based motion capture used in studios to develop games. The team have recorded people's movements in completely new locations — like driving a car — previously out of reach. There is even a video of it in action."
Stop picking your nose and raise your hand if you think this will not be a major boost for pr0n industrie and adult video games?
"oh, oh, oh, Mario, please don't stop, oh, oh, oh!"
A feeling of having made the same mistake before: Deja Foobar
The team have recorded people's movements in completely new locations -- like driving a car -- previously out of reach.
Can't make a clear car out of plastic?
Libertarian Leaning Political Discussion Forum.
Maybe someone who went can dig out their conference DVD and put up the presentation somewhere.
[Fuck Beta]
o0t!
Motion capture by inertial navigation seems a bit expensive, but it's probably affordable for Swiss. Those IDG300's actually increased in price since they came out. Now we're finding all kinds of new uses for inertial navigation. Too bad there aren't many jobs for inertial navigation experts in Web 2.0 crazed Silicon Valley.
Due to vigarious usage, the hand peice has to be replaced.
Again.
So basically, -1 troll/offtopic is really slashdots way of saying "I hate that you thought of something before me."
A few hundred dollars, huh. Are we looking at the beginnings of the real next-gen console(s)?
Porn will take this technology further than everyone else.
meh
This sonic transducer, it is I suppose some kind of audio-vibratory-physio-molecular transport device?
-Peter
This very product has been on the market for years, by Intersense, which also uses accelerometers augmented with ultrasound to prevent drift. It looks like MIT just copied them.
This isn't really new.
http://www.isense.com/products.aspx?id=43&
OMG this might be just what I need to improve my performance in the ring. This would be immensly helpful in critiquing your own performance. My trainer tells me all the time that my leg movements are too wild. Then when I start to look down, he punches me in the face.
Everyday, I bike along a company called Xsense here in Enschede, the Netherlands, which is selling a similar system called Moven as described here as a commercial product.
If I had this prototype I'd be hacking Second Life to stream my movements into the world (and get Linden Labs to stream it out) and everyone would be like "wow, how's he move like that?!" but I guess they'd just think I was a poser.
How we know is more important than what we know.
I'm thinking this will do for the modding and homemade game industry what the video camera did for Youtube.
And this is my Night Elf Mohawk.
That way our boss will be able to tell when we press keys, when we aren't even looking at the screen and even when we flip him off when is walking away. Cool; sign me up (not!).
how long before we can get this thing hooked up to secondlife?
all we need now is some form of feedback and we're all set
all in the interests of science of course
mmmmm science
or you happy to see me?
Shoot, am I the only one that read that as "wearable kilt"?
Long before: http://www.animazoo.com/ I saw these guys give a presentation about two years ago: the processing's almost entirely done with the suit. It's even possible to capture information at a significant distance: a major motorcycle manufacturer used it to work out the applied ergonomics of their more recent sports models, and adapt it so that it would give a 'more optimal ride', I believe the term was.
http://xkcd.com/313/
I think the bigger surprise is that a student from MIT was able to lift so much weight while doing sports!
But they're the worst ping pong players ever.
"-Hunny, you've been standing in the middle of the room waving your arms towards the celling for about 30 minutes now, are you sure you're alright?
-Yeah leave me alone I'm playing Superman!"
You just got troll'd!
ok, better technology, but you still need prep time to slap on the suit and sensors--i.e. prep to look goofy I say.
Someone wake me up when they remove the need to move at all (neural interfaces ftw), so I can lay motionless for the rest of my life.
I think this thing could be huge to professional sports.
:)
You could use this set up to help show athletes how to improve their form, be it in the weight room, on the golf course, ski slope, or any other place where repetitive precision movement is needed and a refinement of form could improve performance.
Heck, even just as a trainer, get your clients to strap this thing on a couple of times (especially those who aren't keen on working out in front of a mirror) to show them their form and how to improve it.
Or combine this with pre-defined motion capture to attempt to train the wearer on how to re-enact the original motions (be it real dancing, DDR, or even 'Ninja Challenge' or what ever that Spike show is!)
For $3k and dropping, the entry fee is so low that there are sure to be people looking make a profit off this system. I'm interested to see what all they come up with
-Rick
"Most people in the U.S. wouldn't know they live in a tyrannical state if it walked up and grabbed their junk." - MyFirs
I find it interesting that in the split-screen presentation, the live video can have a completely different angle than the motion capture video. It looked wrong at first (wait, he's supposed to be facing me...) until I realized that the motion capture video can be rendered from any angle at all. There's absolutely no connection other than time sync between the live and motion capture video.
Then there's the distance. I noticed that on the universal machine the subject was quite a bit farther away in the live video than in the motion capture video. Again, there's no correlation between his physical distance and the distance-from-the-viewer at which he's rendered.
There's probably a practical application of this observation, but I can't think of it at the moment.
That will be all, thank you, more abuse at 11.
This has lots of applications in construction and assessment. I'd like to see all the 3D building structure data loaded into a system and piped to heads-up displays for construction workers, where they and/or structural components are pinpointed by the location sensors. It would give real-time "this thing goes... right... HERE" feedback. The other application I'd love to see is for existing structure assessment: a 'wand' with ultrasound and possibly magnetoresonance to sense pipes, wiring, and structure elements, etc. It wouldn't take long to develop a 3D map of an existing structure, and everything hidden inside the walls.
As for being able to map me dancing? Uh, no thanks.
All pass beyond reach of medicine. None pass beyond the reach of love.
You may have missed the point...
"'The sensors are all off-the-shelf parts,' Adelsberger says, making the system much cheaper than other motion-capture technology. It cost about $3,000 currently, but this could come down to a few hundred dollars, he says, if the sensors are mass-produced."
Forget about buying an already built commercial system, this is a MAKER's dream. Run it on a home-built laptop for extra points.
Just -1, Troll talking to another.
I think the best application would be for use by independant game and animation studios. It's inexpensive, has relatively few components, and doesn't require much space. Perfect for the indie community.
Come on, that was in the last millennium; nobody reads papers from the last millennium anymore, in particular not if they're from the same university. Besides, that's just a product; I mean, products aren't science, are they?</sarcasm>
It seems like the motion that's captured remains in place, with the hips not getting the translation data, so if you were o capture a run, the 3d character would stay in the same spot. This might be useful for creating run cycles for games, but it would practically be useless for anything else, since it would take a long time for an animator to go in and make sure the feet placement is correct and the character isn't sliding around everywhere. Motion capture data is very heavy and fairly hard to adjust once you've captured it.
Don't trust a bull's horn, a doberman's tooth, a runaway horse or me.
I saw this in use over a year ago, then it was a canadian company developing it. Cool stuff, but nothing new.
I can't seem to find anything useful through the wonderful search - I just keep getting back to this article plus a whole lot of unrelated onse - but does anyone else remember an article from several years ago where random students wired up a wearable thing that mapped quake levels and enemies onto the real world you were running around in ? What's new in this one ?
What a depressingly stupid machine.
... is supposed to add what, exactly, to the homemade game industry? Homemade games are not successful on any recognizable scale even when they are made in free development environments for hardware which the entire Western world owns -- after you get past the level of simple Flash game you run SMACK into the content creation barrier and start having to have professional programmers and artists to make a stab at anything. Who is going to write the textures and skeletons, to say nothing of 3D engine (eh, minor detail, that), for those motion capture shots you are taking with your newly commoditized hardware?
Even the semi-pro games are not much to write home about: I used to contribute to the highest activity game on Sourceforge, which at any given time probably had a dozen professional programmers working on it, and it would suffer in comparison to most games from four console generations ago (!) in terms of anything but gameplay. (And that was only fun because we were slavishly implementing a fun boardgame as an online game.)
Help poke pirates in the eyepatch, arr.
This same technology was on the business investment show "Dragons' Den" but invented by a British guy. The investors ended up laughing it out of the room as they couldn't see how it could have any practical application whatsoever. The company's called "Virtual Puppet Ltd" and if you're in the UK (or possibly elsewhere) you can watch their presentation to the dragons here.
The original manufactures of this technology are Animazoo http://www.animazoo.com/ . There have been a number of copy cat systems which claim to be better however on close expection of the data, Animazoo products are far superior. This is due to Animazoo working with House Hold animation companies for the past 10 years and developing a product that the industry wants and not a product that the company thinks the industry wants! Animazoo technology is used by Virtual Puppet as seen in the UK on Dragons Den. This is the Gypsy 6 Torso system intergrated into a puppetering booth.
> Swiss and MIT researchers have developed a wearable kit that will capture your
> every move for mapping onto a virtual character.
Great. A bunch of obese, sedentary-looking fugly pork pies runnin' around with rocket launchers.
(-1: Post disagrees with my already-settled worldview) is not a valid mod option.
This rig would be awesome to take to the local Rock Climbing gym, or even outdoors, to capture the motions involved in rock climbing. Finally something that can capture it in 3D, compared to the current 2D methods used (video cameras arent enough to fully visualize the moves 100% of the time).
Actually, this system was developed at MERL (Mitsubishi Electric Research Laboratories); the MIT and ETH
linkage is that some of the interns working on it were MIT and ETH Zurich.
Check the paper authorships to see this.