The Eye Tribe

Re: Strabismus and EyeTribe

2015-07-07T16:58:45+02:00

Open source project related to this subject:

https://github.com/balancana/GazeMonitor/

Statistics: Posted by skatapuerto — 07 Jul 2015, 16:58

Re: Strabismus and EyeTribe

2015-02-02T12:03:22+02:00

Hi Tom,

It think it might be possible. It really depends on the range of conditions and the severity. Hard for me to make any promises, in theory you should get good data on the "good" eye.

Statistics: Posted by skataMartin — 02 Feb 2015, 12:03

Re: Strabismus and EyeTribe

2015-01-30T06:59:04+02:00

Sorry for late reply.

Martin I don't really care about using it for actual tracking for tasks (clicking stuff on screen etc.)

I just want to be able to detect if the eyes are looking in different directions (at different points) and if so, notify the user. That's it.

What I'm wondering is if this is within the capabilities of the EyeTribe camera and software? I understand it's not completely open source, so you can't do everything.

Statistics: Posted by skatatomtom — 30 Jan 2015, 06:59

Re: Strabismus and EyeTribe

2014-12-11T01:50:21+02:00

Hi tomtom,

It tracks and estimates the gaze direction for both eyes independently but the combination of both eyes typically yields higher accuracy. You can simply choose left or right eye from every sample that the tracker server outputs. There are raw and smoothed values for left, right and both eyes.

Please note that I have very limited data on performance with specific conditions like Intermittent Esotropia. However, if one eye is able to remain fixed/stable during the calibration I see no reason why it shouldn't work (assuming that the condition is not alternating). You might have to tweak the calibration procedure to fit your purpose, we provide source code to get you started.

Statistics: Posted by skataMartin — 11 Dec 2014, 01:50

Re: Strabismus and EyeTribe

2014-12-08T15:01:08+02:00

From what I've seen yes, both eyes are tracked independently and the data is contained in a JSON object returned by the tracker. You could use the tracker to determine this quite easily I'd have thought. Accuracy might be a slight issue however. Currently around the size of a fingertip is the closest your likely to get resolved on a normal screen which may or may not be suitable depending on how bad the deviation is.

Hope this helps

Ian

Statistics: Posted by skataClumpy — 08 Dec 2014, 15:01

Strabismus and EyeTribe

2014-12-06T02:33:26+02:00

Hey all,

So I am interested in the EyeTribe because I want to develop a sort of biofeedback for people with intermittent strabismus (a condition where both eyes don't always look at the same x,y point).

Would it be possible to use the SDK to detect when the eyes aren't looking at the same place on the screen?

For example, the person's right eye could work well and they see with it, but the left eye sometimes looks inward to their nose.
Is it possible to choose which eye to track, and then detect when the left eye turns inward?

In other words, can you track both eyes independently?

This condition probably affects up to 5% of the population, so if this functionality is not possible, it should be implemented I think. Many people won't be able to use EyeTribe without this functionality.

Statistics: Posted by skatatomtom — 06 Dec 2014, 02:33