The EyeHarp

r.erorr · Unread post by **r.erorr** » 29 May 2012, 20:48

It's made for paralyzed people,but I think could be interesting also in this forum
http://theeyeharp.blogspot.com/

nay-seven · Unread post by **nay-seven** » 30 May 2012, 09:48

it's an interesting project, he talk about building a low cost eye tracking device , have you tried..?
btw , I've discover in the article the camera mouse appz , works fine too,( once you have calibrated) except with fader/knob, you'll need to add increment/decrement button, but as we can save a workspace as .exe can be a first approach for people with paralysis, what do you think ?
maybe a new collaborative project here..?

r.erorr · Unread post by **r.erorr** » 30 May 2012, 22:18

about low cost eye tracking device:
No I'm not tried, but I think it depend on physical placement of inexpensive camera + EyeHarp. ..?

about camera mouse app:
Yes, I tried, and works good with inbuilt web cam too. Sometimes it change or lose tracking point, but we can improve that on settings.

abut project:
what exactly do you mean? built instruments on usine specially for "camera mouse" or "eye harp"?

nay-seven · Unread post by **nay-seven** » 30 May 2012, 22:38

here about the DIY : http://eyewriter.org/diy/ seems to be a webcam+ ir leds

about the workspace, yes can be a workspace , with only matrix, button, fader but with increment button and so but the most important is:specially built to be interesting and fun to play with only camera mouse

r.erorr · Unread post by **r.erorr** » 31 May 2012, 01:10

Interesting... I look at EyeWritter PDF, don't know how much time it takes, but seams doable for me. So, let's try!

r.erorr · Unread post by **r.erorr** » 31 May 2012, 04:07

Here an example how to control samples in Usine with Camera Mouse. This is optimizes and improved version of my patch I uploaded previously, but now it is extreme minimalism.

standalone workspace and audio files:
http://www.datafilehost.com/download-0f3aeab7.html

nay-seven · Unread post by **nay-seven** » 31 May 2012, 08:01

It's a really cool workspace but maybe i don't know enough camera mouse settings but I've some issues here :

_here, camera mouse don't like the gain part, it go back to mouse mode each time i choose this part..!
_and btw camera mouse don't let you do click and drag, right..? so you can change the gain..?
_works better for pan, in the way you can change position but no way for loop..?
_changing the area with the mouse wheel...no way with camera mouse..?
_ Drag & drop samples..how do you do this with camera mouse ..?

Sorry for this long issue list but i suppose all must be done easily and with no external help if you want the user feel independent ..?

last thought : not sure the sample choice is a good idea as the people with paralysis live half of their time in hospital..? maybe you need to provide a more positive trip..?

r.erorr · Unread post by **r.erorr** » 31 May 2012, 10:30

full controll? of course no way. I just use main panel in realtime.
But you are absolutely right about independence, then we must choose most universal system like notes, at least on start.
I will tray to make something to night.

martignasse · Unread post by **martignasse** » 31 May 2012, 14:06

hello all,

for infos, apart from my usine work, i'm on a project where we are building an eye tracker prototype derived from the eye writer 1.0 (mounted on glass, not the 2.0 mounted on the pc)

well, the 1.0 is really cheap to construct but you'll need some good manual aptitudes
the biggest limitation of the 1.0 version (for us, non handicapped) is that you shouldn't move the head at all or you'r calibration is down

first result are very impressive, the mouse cursor literally go where you look at... it's fun and strange

for now, i'm fighting with the calibration and precision because the cursor is a bit too jerky for a comfortable use

nay-seven · Unread post by **nay-seven** » 31 May 2012, 15:03

Whaoo, cool news
and nice coincidence ! is the click work the same way ( Dwell time ) ?

r.erorr · Unread post by **r.erorr** » 31 May 2012, 16:04

martignasse wrote:hello all,
you shouldn't move the head at all or you'r calibration is down

starnge, because the same problem is for v.2 too, isn't it? therefore this fun muffler

Can you tell more something about cause - why the calibration is lost in v.1 ? Is it really because of head movements or because of unstable construction [glasses + webcam].

martignasse · Unread post by **martignasse** » 31 May 2012, 16:17

don't know about the v2.0,
i'm focused on the v1.0 and the final user has the looked-in syndrom

head movement is the first calibration killer
the construction can be a part of the problem too (jaws movements can move the glass)

but even if you minimise/resolve these problems, the soft calibration is very complex
the appli provided (coded with openframeworks) isn't very clear (mean rapid coding) and i'm suspecting some unfinished part in the calibration process.

r.erorr · Unread post by **r.erorr** » 31 May 2012, 16:19

Maybe we shout start not with glasses, but head or body holder chair?

r.erorr · Unread post by **r.erorr** » 31 May 2012, 16:24

Jus one question for future. Is it possible that soft can detect eyes focus changes? when we focus on smaller or large spots

iococoi · Unread post by **iococoi** » 31 May 2012, 17:13

hi there!

did some research on eye tracking a while ago...found this rather helpful list of currently available eye trackers and other gaze interaction related stuff..

http://www.cogain.org/wiki/Eye_Trackers

~

r.erorr · Unread post by **r.erorr** » 31 May 2012, 20:57

Thanx, will check ...

Welcome to %s forums

The EyeHarp

Who is online

Login • Register