random idea Re: "scientific" gesture / movement research ?

This is definitely an aside to the ongoing thread. But may be
applicable for next memory-place experiments (not as sophisticated
as UI device)... Just something I've been loosely imagining.....
Without yet thinking through the tech process or synthesis....

A couple of references first:

Depth and gesture mapping / tracking of a participant:
http://jmpelletier.com/freenect/

I was thinking that this environmental installation project
might offer some intriguing possibilities:
http://kinecthacks.net/spinny-glowy-foil-in-a-kinected-bunker/
In particular, notice the white spindly sculptural things (you'll see
them toward end of the video) hanging from the ceiling.

I was imaginging, for example, what would happen if we built
the equivelant of the white 'glowy foil' sculptures - but made
them out of strips of IR LEDs. Then use a matched IR detector
apparatus for triggering vibratory feedback (EG:
http://www.radioshack.com/product/index.jsp?productId=2049723)
pretty much as our last device did. Only in this case, it would
be the overall temporal experience of detection, that might add
up to the participant's recognition of distinct shapes. They might
undergo a slow process of meading one detection from another,
as they explore the space.. Slowly developing a physical
relationship with the way they share the space with the
sculptures (known to them only as on/off vibrational feedback).

Pelletier's note (first link above) is that more than one kinect
can be in use at once. It might be messy to work the bugs
out (interference wise) but; one kinect program could be utilized
to trace participant movement in relation to the IR sculptures,
and another could be employed for retreival of depth and gesture
informations... If that's the sort of information we hope to
retrieve.... I haven't thought out the means of integrating info,
at this point at all....

Also, I was thinking about the issue of participant's bumping
into the sculptures.... So.... What if we used transparent cloths
as dividers that were situated between participants and the
sculptures? The IR sculptures could be encircled with these
cloths, hung from the ceiling to floor... Light-weight and soft,
and distanced enough from the sculptures to serve as an
adequate indicator (to the participant) that they should move
in another direction. This instruction could be provided in
advance of the participant's blind-folded exploration (whenever
they feel the cloth against their skin, they should stop and
turn away from it). Would simplify process, avoid collisions...

Could use battery power, rather than cabling, for the IR
sculptures.... Clearing out and freeing up space for participant's
to move in.... With exception of the cloth veils, of course...

?

x patricia


----- Original Message -----

From: Sha Xin Wei

Sent: 05/31/11 06:41 AM

To: Michael Fortin

Subject: Re: "scientific" gesture / movement research ?

I suggested that some one open up a wii mote  and re-assemle the parts into a suitable form factor.  we still need a visible light photocell though, and cant use  line of sight, so that solution is also clunky...
 
We need a good versatile engineer to own this project and work with the MP group.
 
And strategically in June I'd like to define a non-hobbyist grant to NSF/NSERC/FQRNT parallel to an MP grant to do a gesture/movement tracking research project that meets different interests around the TML -- MP, Adrian (+Wessel), MM, Satinder.   I'll propose this  to my EU colleague as well.
 
Xin Wei
 

On 2011-05-30, at 6:43 PM, Michael Fortin wrote:

This might be a jaded comment.... 
 
I'll call it an advanced WiiMote (WiiMote just tracks the x-y-rotation, they have some idea of angle and distance to the display which the WiiMote doesn't have).   (Morgan -- WiiMote has vibrotactile feedback)
 
Speaking of odd remotes, there's this unrelated interesting toy; http://www.thinkgeek.com/interests/dads/cf9b/
 
Cheers,
~Michael();


On Mon, May 30, 2011 at 03:55, Sha Xin Wei wrote:
Hi Adrian, and scientific researchers,
 
Raising the stakes and thinking ahead to more robust and precise instrumentation, here's the
 
NaviScribe 6-DOF 3D wand by Electronic Scripting Products, Inc. (ESPi) in Palo Alto
 
The exclusive patent describes a  6 DOF, x,y,z + euler angles    The company's founded  by a physicist friend from Stanford: Marek Alboszta.  Not productized yet.  "Commercial" co-development would require O(100K) USD.  I've not discussed how to enter into actual relation with this company, but we could perhaps work out a deal.  This would make sense in a real NSERC/NSF  co-development grant.
 
Shall we think about this in context of a scientific gesture research proposal, along with high FPS cameras, and EONYX etc?  Let's discuss this in June.
 
Xin Wei
 
 
 
 

On 2011-05-20, at 9:32 PM, Marek Alboszta wrote:

Hello Xin Wei,
 
We can definitely do everything you ask (briefly - up to 100 Hz and better with all degrees of freedom (6DOF) reported in compact stream (right now not compressed), requires at most 120 MIPS to do everything (during periods of a lot of activity) - unit is small so can be in a ring or glasses or headgear or whatever you choose - we give you intervals so you can compute your derivatives, resolution in 3D space is considerably better than 1 cm (in plane it's down to 0.2 mm and better)).  I can't do wireless unless somebody gives me money to properly design a wireless beta unit (it is not a problem of technology but pure resources).  
...

Is your party ready to pay for this work ...?  If not then we should reschedule for when they are ready to commit resources for technology development (or if they/your side wants to do the work).  Anyway, we can talk about it if the allocation of resources is a given - let me know.
 
warm greetings,
<pastedGraphic.jpg>

_____________________
Marek Alboszta
 

On May 20, 2011, at 10:08 AM, Sha Xin Wei wrote:

Hi Marek,
 
For a memory & place experiment, we would like to give people a wand that they can hold that can report position, euler angles, and their time derivatives.  Ideally at better than 30 Hz for the entire 12-vector.
 
We need it wireless, range of say 10m suffices.
 
Spatial resolution is important, for tracking "pointing" at virtual objects that people infer by indirectly mapping position & angle to a vibration motor that will be embedded somewhere on their body.   I expect any pen-based input device has more than adequate time-space resolution.
 
We would also like to be able to have a "wand" small enough to fit anywhere attached to the body in some not too obstrusive way.
 
We can write our own code to parse the data if you tell us the format coming in some standard protocol, serial or ethernet/port stream.
 
The person may be free to wander around the room and point in any direction whatsoever.
Does the wand needs to see an IR array in "front" ie be constrained to a half-sphere, or can it be pointed in any direction provided a set of IR beacons ...
 
Cheers,
Xin Wei
 

 


...............................................................................................
Because the essence of technology is nothing technological, essential reflection upon technology and decisive confrontation with it must happen in a realm that is, on the one hand, akin to the essence of technology and, on the other, fundamentally different from it. Such a realm is art. But certainly only if reflection upon art, for its part, does not shut its eyes to the constellation of truth, concerning which we are questioning." - Heidegger