×

Welcome to the Slashdot Beta site -- learn more here. Use the link in the footer or click here to return to the Classic version of Slashdot.

Thank you!

Before you choose to head back to the Classic look of the site, we'd appreciate it if you share your thoughts on the Beta; your feedback is what drives our ongoing development.

Beta is different and we value you taking the time to try it out. Please take a look at the changes we've made in Beta and  learn more about it. Thanks for reading, and for making the site better!

Breaking Motion Capture Out of the Studio

Unknown Lamer posted more than 2 years ago | from the in-disney-land-actor-captures-motion dept.

Input Devices 39

Fnord666 writes with a CMU press release. From the article: "Traditional motion capture techniques use cameras to meticulously record the movements of actors inside studios, enabling those movements to be translated into digital models. But by turning the cameras around — mounting almost two dozen, outward-facing cameras on the actors themselves — scientists at Disney Research Pittsburgh, and Carnegie Mellon University have shown that motion capture can occur almost anywhere — in natural environments, over large areas, and outdoors."

cancel ×
This is a preview of your comment

No Comment Title Entered

Anonymous Coward 1 minute ago

No Comment Entered

39 comments

MS Kinect (1)

jweller13 (1148823) | more than 2 years ago | (#37036566)

I have the xbox kinect and I would think it would be a great tool for this. From what i understand many of our nerds brethren have a disassembled kinect's and retooled and programmed them with all kinds of funcky capabilities.

Re:MS Kinect (1)

MrBandersnatch (544818) | more than 2 years ago | (#37036928)

The problem with Kinect is the limited minimum range (50-80cm) at which depth information can be extracted; while visual systems are affordable and in the price range of researchers to study, in the long term electromyography probably will be the long term technology of choice.

Re:MS Kinect (1)

RoFLKOPTr (1294290) | more than 2 years ago | (#37081546)

in the long term electromyography probably will be the long term technology of choice.

I'm sorry, but do you have a license from the Department of Redundancy Department? I'm going to have to see some documentation.

Inevitable (1)

Anonymous Coward | more than 2 years ago | (#37036582)

In Soviet Russia, motion captures you!

In motion studio.... (0)

Anonymous Coward | more than 2 years ago | (#37037256)

...YOU capture Soviet Russia!

Just a matter of compute power (1)

Isaac-1 (233099) | more than 2 years ago | (#37036592)

This sounds obvious, but when you look at it like many things these day it is a matter of compute power, particularly having enough at a reasonable price to make it practical.

Re:Just a matter of compute power (1)

GoodNewsJimDotCom (2244874) | more than 2 years ago | (#37036744)

Actually the software that figures stuff out is remarkably complex. I wonder how long until someone makes a game where you wear VR Goggles and carry a nerf sword, and you're out in a field killing orcs by swinging the nerf sword around.

Re:Just a matter of compute power (1)

ColdWetDog (752185) | more than 2 years ago | (#37037184)

Actually the software that figures stuff out is remarkably complex. I wonder how long until someone makes a game where you wear VR Goggles and carry a nerf sword, and you're out in a field killing orcs by swinging the nerf sword around.

And how long after that until there is an uptick in arrests for bizarre, aggressive behavior in usually introverted males?

Re:Just a matter of compute power (2)

nomel (244635) | more than 2 years ago | (#37037452)

Well, it's mathematically complex, but not really complex in a code size or if-then count sort of way.

Here are two videos (1 [youtube.com], 2 [youtube.com]) from 2007 showing the jist of it (including some games that modify reality) done in *realtime* using a laptop GPU card.

It's basically...find a bunch of "features" in a scene (corners in some texture, circles, etc), then look at how these features move in the scene as the camera moves. Each feature is in a fixed point in space, so, their movements are caused by the camera moving. Using basic multi-view type math from the multiple views brought on by the movement and each frame of the video, you can "solve" the cameras position, rotation, AND all of the feature positions in 3D space using a big giant matrix created by all of those 2d image feature positions. If you've taken first year of linear algebra, and you covered singular value decomposition, you'll understand probably 80% of the math in the research papers (they'll find crazy efficient ways to do things...though.

Some complexity comes from disregarding moving objects in your scene...but that's mostly a "throw out data with a wtf factor greater than x" type statistics problem.

GPU's are very good at finding features, which is why that video from way back in 2007 is possible...that's around the time that shader programs with reasonable complexity could be made (like more than three texture lookups per pixel). The lack of complexity is also proven by the huge number of flash applications a few years ago that used this with your webcam, and the existence of the Nintendo DS AVR games, and all the cellphone apps that use it.

This isn't bleeding edge or all that complex...but it is a very cool. It's more of the standard, "the hardware and software is cheap enough to do it now".

If you're interested in this stuff, you could get something reasonable going using OpenCV pretty quickly...most of the feature tracking routines are built in.

On a side note...look at this cool shit we could be working on if people would stop messing with f'n web browsers/apps...

Not sure there's an advantage (0)

Anonymous Coward | more than 2 years ago | (#37036610)

The article points out that it's slow and inaccurate. The cameras are bulky and would get in the way of the actors. The advantages seem to be outweighed by the limitations.

Mocap is highly overrated anyhow. The best animation I've seen lately was in Rango and although they used actors for reference it was all hand animated. Straight Mocap always looks fake it needs a lot of tweaking to look right. It's an interesting reference but it's no replacement for a talented animator.

Re:Not sure there's an advantage (4, Interesting)

Isaac-1 (233099) | more than 2 years ago | (#37036702)

Just wait for version 2.0 where the cameras are smaller than the size of a coin and everything is neatly sewn into a spandex motion capture suit. Think of potential over real terrain, rock climbing, tree climbing, etc.

Re:Not sure there's an advantage (1)

nomel (244635) | more than 2 years ago | (#37037544)

I'm assuming the cameras are big because they hold a get nice big, high sensitivity, sensors and probably low F-number lens. Motion blur on something like this would make things MUCH more computationally difficult compared to snapping a clean picture and figuring out which way the camera is now facing. Motion blur can be estimated [google.com], but you wouldn't want to do that every frame...

Re:Not sure there's an advantage (1)

RoFLKOPTr (1294290) | more than 2 years ago | (#37081570)

It's an interesting reference but it's no replacement for a talented animator.

Talented animators are very good at what they do, but you're wrong. Motion capture is the most accurate method of character animation in existance... it's a matter of having a complex enough rig and a good enough actor. If you have those two things, then you have perfect animation every time. And it's cheaper and less time-consuming too. Something tells me that you're an animator that's afraid he might be out of a job soon.

Motion Sensors? (1)

StarKiller53861 (2251214) | more than 2 years ago | (#37036620)

Is it possible to get the same result using a series of motion sensors (like the Nintendo Wii uses)? If not, then what is the cheapest/easiest way to capture real-life movement (and apply it to 3D models, for example)?

Re:Motion Sensors? (0)

Anonymous Coward | more than 2 years ago | (#37036858)

Errors creep up in IMU measurements because it's an open loop system but using cameras, each point can compute it's location absolutely in the environment. If there's an error in any reading, subsequent readings aren't poisoned. Plus, (good) IMU's can be expensive compared to cameras. It's also pretty simple to just record the video from each camera and post-process it for the location information. An IMU system would need to be real-time recording the vector movement of every point.

Also, Wii-mote's to know their direction do rely on a camera. It see's the four IR LED's on the "sensor bar" for it's orientation information. If they can't see it, they don't know which way they're facing. And since now you're putting a camera anyways, how about we only put cameras?

Interesting approach, but also... (4, Interesting)

luckymutt (996573) | more than 2 years ago | (#37036696)

Interesting approach. I would have loved to have seen the presentation, but also today at SIGGRAPH, Autodesk had a presentation by Kevin Wittkopf [autodesk.com] and Geoff Richardson [autodesk.com] showing off using the Kinect in production with Maya. Why bother with a suit and a bunch of cameras when Kinect rigs and only going to be getting better (and waaaayyy cheaper) Autodesk is streaming their presentations [autodesk.com] from the exhibit hall. I'm not sure if they are going to do that prese again tomorrow or Thursday, though. (please don't everyone check it out at the same time...I'm trying to watch it)

Re:Interesting approach, but also... (1)

luckymutt (996573) | more than 2 years ago | (#37036726)

hell, it looks like they are giving it again right now. (they've been way off of their schedule all day today)

Re:Interesting approach, but also... (1)

Anonymous Coward | more than 2 years ago | (#37036778)

because the kinect won't work outside or in large areas.

So how long .... (2)

gstoddart (321705) | more than 2 years ago | (#37036786)

So how long before the inevitable application of "Rule 34" to this?

Let's face it, the porn industry has been a leader in adopting technologies for some time now.

Re:So how long .... (1)

operator_error (1363139) | more than 2 years ago | (#37036902)

Oblig XKCD to explain rule #34 in-a-click: https://www.xkcd.com/305 [xkcd.com], for those that provide relevant cultural references without one.

Re:So how long .... (0)

Anonymous Coward | more than 2 years ago | (#37039234)

Couldn't you bother to find the porn version of that one?

BTW: XKCD totally should a Rule-34-ed version on April 1st next year. Fuck SFW! ^^

Yada yada yada, dl link? (1)

Ceriel Nosforit (682174) | more than 2 years ago | (#37037084)

Blender Tomato Branch

http://www.vimeo.com/26420002 [vimeo.com]

Latest build
http://graphicall.org/?keywords=tomato [graphicall.org]

You can find the devs on #blendercoders @ freenode.

Re:Yada yada yada, dl link? (1)

Ceriel Nosforit (682174) | more than 2 years ago | (#37037108)

Re:Yada yada yada, dl link? (1)

Ceriel Nosforit (682174) | more than 2 years ago | (#37037280)

I mixed up motion capture and camera tracking. Sorry.

Although... it seems that if the Tomato branch system captures points on a subject from two angles it could reconstruct motion data since that is sort of what it does.

Re:Yada yada yada, dl link? (0)

Anonymous Coward | more than 2 years ago | (#37041916)

Yes. I saw once a video of an asian guy showing how you can make basic motion capture with the Tomato branch. He shot himself moving an arm with some colored dots on it, with the camera still, and then he tracked the dots and used that data to superimpose a robotic arm on top of the original video.

Is this all the same problem? (1)

malignant_minded (884324) | more than 2 years ago | (#37037104)

Sounds like we are mixing two problems and the video shows a problem that seems to me not necessary in solving. I'm no expert or even n00b on the subject but I never really understood why you would need a camera and florescent dots on a person to record motion. Why can't you just have sensors that measure the movement like my phone, and you could do this anywhere. The video recording of dots seems like the old school poor mans way of recording points and thereby recording motion. Now if you want a three dimensional scene to be recorded along with the motion then yes it makes sense to use a camera but I don't see how a three dimensional scene could have been created in the studio anyway by recording dots. Now if we are talking about the matrix like freeze while moving around a three dimensional object isn't that a different problem? From what I recall this is done by recording with several cameras but is motion really involved? Now perhaps both of these problems can be solved now using this new method since you now have a three dimensional reference for the scene but doesn't the data need to be collected separately? I mean if I just stand still and wanted it to be a shot of me standing on a hill while the cameras do that helicopter circle pass shot that is over done what three dimensional model have I recorded for the scene since I have not moved?

needs more refinement (1)

slshwtw (1903272) | more than 2 years ago | (#37037574)

You can look at the video here [disneyresearch.com].

The motion-capture of the guy swinging from the monkey-bars looks somewhat realistic until you realize that his hands are swinging around in the rendering when they should be (and are) stationary on the actor.

Should be interesting when the technology matures, though.

limited use (1)

nEoN nOoDlE (27594) | more than 2 years ago | (#37037908)

The accuracy on this is pretty bad, as you can see from the comparison to the Vicon system, so this has very limited use in production, and it seems like it's trying to solve a non-existent problem. It doesn't get rid of the cumbersome suit from motion capture, so you can't really get capture data while performing in front of a camera, and there are already solutions [ipisoft.com] on the market that allow you to create a motion capture setup outside using regular cameras and get motion capture data without any tracking markers. The only new thing this brings to the table is having a really wide range of capture, as the test with the person walking along the curved path shows, but the practical application for something like that is limited. Attaching cameras and trying to create tracks off of the movement of the background just seems a really backwards way of doing this kind of stuff.

Re:limited use (1)

iceaxe (18903) | more than 2 years ago | (#37038280)

The accuracy on this is pretty bad, as you can see from the comparison to the Vicon system, so this has very limited use in production

This is a step along the way toward something useful in production, not a final product.

[...]the practical application for something like that is limited. Attaching cameras and trying to create tracks off of the movement of the background just seems a really backwards way of doing this kind of stuff.

The application is in capturing motion in circumstances where the current way of doing things does not work well. Larger ranges and areas, confined spaces where cameras would be obstructed, that sort of thing. It doesn't replace current techniques, most likely. Instead, it adds new options - once it's accurate enough. Time will tell.

Lord of the Rings (1)

Neil Boekend (1854906) | more than 2 years ago | (#37041194)

I can remember they did something similar for Gollum in the Lord of the Rings. For the last scene they filmed for the Return of the King they didn't have the time to capture Andy Serkis after filming so they stuck him in the mocap suit and captured him during the filming of Elija Wood and Sean Astin. This saved a load of time in postproduction, and due to the interaction Andy Serkis had to be there anyway.
Check for New Comments
Slashdot Account

Need an Account?

Forgot your password?

Don't worry, we never post anything without your permission.

Submission Text Formatting Tips

We support a small subset of HTML, namely these tags:

  • b
  • i
  • p
  • br
  • a
  • ol
  • ul
  • li
  • dl
  • dt
  • dd
  • em
  • strong
  • tt
  • blockquote
  • div
  • quote
  • ecode

"ecode" can be used for code snippets, for example:

<ecode>    while(1) { do_something(); } </ecode>
Sign up for Slashdot Newsletters
Create a Slashdot Account

Loading...