The Eye Tribe

General Discussion • Jumpy pupil diameter

2016-11-25T12:16:20+02:00

Recently I discovered that the pupil diameter signal can contain artificial "jumps" which are well above the noise level. I guess that these "jumps" are caused by adaptive processes in the camera (automatic exposure control) and/or the eyetracking algorithm (adaptive threshold for pupil edge detection). The amplitude of the jumps seems to be very similar in one recording but can be different across recordings (different subjects, different conditions). The amplitude of these "jumps", I guess, depends on different individual pupil-to-iris contrasts and different levels of focus (also depending on glasses vs. contact lenses vs. no correction, even at the same eye-to-camera distance).
Obviously, having the exposure and threshold controlled adaptively is a good thing, but I wonder if we could have a "tracker" attribute in the API which allows us to freeze/unfreeze the currently used exposure level and threshold in order to get an artifact-free pupil diameter signal when this signal is of particular interest.

.
Pupil diameter of one eye over time. X:500ms/div, Y:2.5pix/div (where "pix" is the unit of "psize" as returned by the TET API).

Statistics: Posted by skataqx7711 — 25 Nov 2016, 12:16

General Discussion • Re: Eye Tribe Pro Release Date

2016-11-06T14:38:41+02:00

I am planning to do a project using the device in december. Did any of you guys get your device?

Statistics: Posted by skataajith.murugaian@gmail.com — 06 Nov 2016, 14:38

General Discussion • Eye Tribe Connected but not working

2016-10-24T11:26:32+02:00

Hi everyone, I am using eye tribe but suddenly can not detected. It gives a message that your device has been connected but not working. Anyone who have an idea on how to resolve this problem, Actually, This is the second device I experienced the same problem. Maybe the Eye Tribe Company should reconsider this, the device is quite cheap but not durable at all. Its annoying, in the middle of experiment the device just stop working suddenly!! #feelingconfused

Statistics: Posted by skatakimarionizetha@gmail.com — 24 Oct 2016, 11:26

General Discussion • Re: Eye Tribe Pro Release Date

2016-10-21T06:13:01+02:00

+1 Tried the contact form but haven't heard anything. Waiting 8+ months now, how about you?

Statistics: Posted by skatamattfeusner@yahoo.com — 21 Oct 2016, 06:13

General Discussion • Scan path

2016-10-20T07:00:52+02:00

hii, can you help me to find scan path or fixation points from gaze coordinates, I am using the eye tribe to monitor a game environment.

Statistics: Posted by skatavenusudha13 — 20 Oct 2016, 07:00

General Discussion • Re: Eye Tribe Pro Release Date

2016-09-29T11:10:22+02:00

Same question here !

Statistics: Posted by skataOlivier — 29 Sep 2016, 11:10

General Discussion • Re: Eye Tribe Pro Release Date

2016-09-28T10:26:10+02:00

Hi,
I have a similar question: I have planned some experiments late 2016 / early 2017, and I wonder if the new device will be available at that time?
Sincerely,
Francis

Statistics: Posted by skataFrancis.Jambon@imag.fr — 28 Sep 2016, 10:26

General Discussion • Eye Tribe Pro Release Date

2016-09-23T15:36:53+02:00

Hi,
Could you give some information about Eye Tribe Pro release date ?

We ordered on may but still we don't know the release date.

Could you explain the delay problem ?

Regards,

Emrah

Statistics: Posted by skataEmrah — 23 Sep 2016, 15:36

General Discussion • Re: Pupil Diameter

2016-09-18T11:37:38+02:00

Hi,

I have a question about pupil size. I have read from the forum that pupil sizes are recorded as pixel by Eye Tribe eye tracker.
But, I want to convert it to mm.
For the conversion is it enough to multiply pixel value by 0.264583 ? (because of 1 px = 0.264583 mm).
Or, is there another formula for calculation, if the person is 70 cm far from eye tracker and monitor.

Can you please help,
Many Thanks.

Statistics: Posted by skatagoncagmen — 18 Sep 2016, 11:37

General Discussion • Position of tracker in car

2016-09-12T13:25:04+02:00

Hi,

we are working on a science project about eyetracking in the cockpit of a car to enhance the user experience.
The test person is sitting on the driver seat and looking at the display in center console of the cockpit, behind the gear shift lever. So far we were only able to place the eyetracker below the display to get a working calibration.

Is there any way to put the tracker above the display and still get it calibrated?

Statistics: Posted by skatat.hofmann@hs-osnabrueck.de — 12 Sep 2016, 13:25

General Discussion • Re: pupil size

2016-09-10T09:44:11+02:00

pupil size- returned through pixels the matrix capturecamera / used https://www.leopardimaging.com/USB_3_3.html/

Statistics: Posted by skatabigpepperlo@gmail.com — 10 Sep 2016, 09:44

General Discussion • pupil size

2016-09-09T14:26:58+02:00

Hi,

I have a question about pupil size. I have read from the forum that pupil sizes are recorded as pixel.
But, I want to convert it to mm.
For the conversion is it enough to multiply it by 0.264583 ? (1 px = 0.264583 mm).
Or, is there another formula for calculation, if the person is 70 cm far from eye tracker and monitor.

Can you please help,
Many Thanks.

Statistics: Posted by skatagoncagmen — 09 Sep 2016, 14:26

General Discussion • Going from 9 to 16 calibration points stops jittery tracking

2016-08-07T22:59:40+02:00

A couple days ago, I was noticing that the tracking was getting a bit spotty.

I then noticed that I had my calibration points set at 9.

I usually have it at the maximum points, but during some time, it was reset.

It's possible that the tracking recovery is coincidental, and 9 points is usually adequate.

I wonder what would happen if you could use more points, but I'm guessing that there are diminishing returns.

Statistics: Posted by skataJeffKang — 07 Aug 2016, 22:59

General Discussion • Re: .net framework

2016-07-25T06:21:54+02:00

I also have the same problem. I am using c#, I downloaded a sample code from Github and i tried to run it in visual studio environment, of course i have already installed .net framework 3.5 as they recommended in the website but when i run the sample code i found out that some library is missing. any way with an idea on how to solve this problem please

Statistics: Posted by skatakimarionizetha@gmail.com — 25 Jul 2016, 06:21

General Discussion • Re: Eyetribe - Head Movements

2016-07-22T10:56:20+02:00

Martin wrote:
Hi DP,

The first development-kit device does not account for head movements along the Z-axis (depth) from the device.

We know that's been a big limitation so we've worked hard on it over the last two years. New device resolves this issue. Other improvements for the new device is larger screens (up to 27"), higher frame-rate (+75), pupil size in millimeters, and more robust USB connectivity.

Hi Martin,

Is there any visual comparison available to see the effectiveness of the new head compensation? We are trying to implement our own, but it may be worth to give the 2nd gen evaluation kit a go, depending on the improved accuracy.

Statistics: Posted by skataCheveoner — 22 Jul 2016, 10:56