forums.naturalpoint.com

I'm very much a newbie when it comes to motion capture, so please forgive me if this question is considered... simple.

Is it possible to use a one camera setup with the actor facing the camera (two to three feet distant) and using smaller markers to capture facial animation? It seems possible, but since I'm not familiar with mo-cap and this particular system (or ANY system) I'm not sure.

I'm not sure about other systems, I think some of the highr end systems can, but we will be releasing a version of the software which will be able to do facial capture, in the future.

DSW:

All that I've seen requires I think at least 2-3 cameras... has to do with triangulating a coordinate in space. There was a paper I read somewhere on the internet though, that some japanese researchers were using mirrors to accomplish this...with a single camera...

they had the actor sit in front of a triptich mirror... where you could see the actor face full frontal, while on the side mirrors you could see the face in a 3/4 position... they then took that information and faked the software to think they had 3 different views... I don't remember how they did it though...

ok... enough of my ramble.... grin

Bryan

Okay, now that's just WAY too fast for a response!

I know you obviously can't state an exact release date, but any idea of what "in the future" actually means? Would it be within the Arena 1.x cycle or much later? Within a year or much farther?

In any event it would not discourage our purchase. It's an incredible system!


Thanks

It is something that we are looking into, for release after rigid body, but it might be pushed out slightly.....

http://www.naturalpoint.com/optitrack/p ... rison.html

Sounds good. Looking forward to the upcoming releases!

Thanks for the info.

Interesting idea. I was just hoping that the basic capture would have at least sufficient data to work with. Looks like we'll be getting it - just not sure when. Fortunately we're not is a great rush.

Thanks for the info Bryan.

Dave Williams

Well if the idea is to do facial capture with one camera to keep the system lower cost, that's one thing...but with this base and having four or six cameras...you'd just use them.

I would personally find facial data capture extremely valuable and a high selling point.

::::: Opera :::::

Hello:

Yes, we also value facial mocap, and are working on solutions for it. We plan to have software that will use your existing V100 cameras for facial, so you can repurpose them.

Hey it's about 5 months after this last post and I'm curious how this is coming along.

1. Will facial mocap be coming out this year?

2. Just to be crystal clear, how exactly does facial animations fit in with the full body? Would the User have to:
. Capture the facial animation in one file (BVH)
* capture the full body motion (another BVH)
* Apply the Full body motion to the Model
* Apply the Facial animation to the same model (on a designated frame)

Is that the workflow in a nutshell?

This is very interesting.
Thanks