Learning visualization...

Hi,

I'm a graduate student at Purdue University studying motion capture and gesture recognition and along with my love for POI i would like to focus my study towards visualization and the learning of new POI or Staff techniques, or maybe about anything pertaining to difficult gesture explanation. I've seen many posts where people get mixed up on terminology and descriptions and this is where video comes into play very nicely but video doesn't always show the angle one wants, etc.. Anyways, i would like to know what the major problems are when trying to describe a move, learn a technique and also what would you (as a viewer) like to be able to see/know/visualize when learning a move/technique.

Any suggestion is greatly appreciated. Keep in mind that this topic is open to any technology concerning 3D animation, video, motion capture, and artificial intelligence.

Thanks in advance,

full 3D + 360' view with zoom and slow motion and a 1st person view.

The hardest part of visualizing a new move is being able to see it through your own eyes since you have never been able to see it from the spinners perspective.

some verbal exsplanation of the move with a short rundown of critical hints and tips where applicable.
ie. when doing a hyperloop try to make the tangle point at the middle of each string.

agreed a full 3D representation would make it so much easyer.

if it was a 3D animated model u could easily slow the motion down move the camera in the 3D space to see it from any angle u thought would help. it would need to show detail of wrist/hand movement in perticular.

and use motion capture to animate it. i saw a program that showed poi patterns using maths and sincurves etc. the moves were far too perfect (some times the poi arnt in perfect planes), it made it hard to learn from.

u could have 1 animation that has a voice over and directs the camera to key parts of the move (essentially a short movie) then open a 3D model for you to "explore" showing the move

Thanks a lot. I never would've thought of a voice directed camera which would be GREAT. And yes, the source of the motion would be in fact motion capture, not procedural animation such as sin curves, etc... The lab i work in has cameras that run at 200fps but for the fast poi motion it needs to be quite a bit higher so we are looking into getting new cameras. Please let me know if you have any more ideas, these are great.

Thanks,

oh, another question. As DarkFyre pointed out "Critical Hits". Do you think it would be beneficial for each motion to have a versioning approach so one person could "tag" information to a clip (at that point in the video/ 3d space) and others could see these tags (information) and how the move is viewed by others?

Sadly I think the tactile is most important in learning something. Most all of us could easily spin blindfolded if we wanted, it's the *feel* of the poi that will let you know if they're off first..

That said, I can't think of any way to replicate that, so I would say the first person visualizations would be next. Moves look very different from watching someone else, to actually being *inside* the move.

Yes i agree, tactile is the best by far... no substitute. But not everyone is able to buy plane tickets and meet up with each other to learn new moves and of course thats where these message boards, videos, etc.. come in at. Thanks for you suggestion ... i will keep it in mind. The first person perspective could have some interesting bloopers when learning new tricks..haha.