I need to buy an eye tracking device to use in an experiment in which I will have to track:
a) the gaze movements of someone tracking an object moving on the screen
b) the saccades between text segments
That's a tough call for 400$, given that you want to have a good spatio-temporal resolution for saccade detection. Temporal resolution (e.g. sampling frequency; fS) will cost you a LOT, but if you are not interested in the saccades but more the fixation pattern (which is the case for reading as I understood it), you could save a lot of money by choosing a lower fS. The lowest would be from common cameras at 50Hz (but that will also introduce some spatial error, since you will not detect the full saccade for short saccades).
You could use the PUPIL project (although I do not have any experience with that). They sell a monocular setup for 380$. Unfortunately, the camera runs at 30Hz and I have not found info on the spatial resolution of the setup.
Alternatively, you could built one yourself and use their freely available software to detect the eye movements, which would trade some money for your time, but would get you easily down to around 100$.
You could also consider to use a rather old-school technique: electrooculography. If you have connections to your local university clinic, you'll get the equipment (or at least the electrodes) quite cheap.
I attached a PDF on eye tracking methods (which gives a nice summary on EOG).
Greetings David
http://pupil-labs.com/pupil/
Article Eye movement recordings: methods
See the eye tribe-Tracker, a 60hz remote (99 USD). In combination with this device you can use a freeware for stimulus presentation and recording.
I am using this combination: Eyetribe+Ogama freeware
It is ok for picture and big AOI based research.
http://theeyetribe.com/
http://ogama.net/
If you want to build your own system, you can try kinect camera which is very goog at object localization, and cheap (
A further question is: only the eye-tracking (hardware) device or also the software? Hardware starts at 30€ for two simple webcams, however, in this case you need a software which does allow you the integration of generic hardware devices. Maybe there are freeware solutions, but I don't know them and I also have no idea concerning the quality.
Dear David, Matthias, Xuelian and Andreas, your answers helped me quite a lot. I bought the "eye tribe" device. I will soon give you feedback on your fantastic answers..
I'm trying to run and modify the EyeTribe-Toolbox-for-Matlab, calibration seems to be ok. Thanks to Edwin Dalmaijer for his job.
https://github.com/esdalmaijer/EyeTribe-Toolbox-for-Matlab
Hi Nelson,
I just wanted to add some more information here in case anyone ends up on this thread (and now that The Eye Tribe has been bought by Google, unfortunately taking those eye trackers off the market).
There are a few options I know of that are available at relatively low price points (e.g. Gazepoint GP3, Tobii Nano (laptop only), EyeTech VT3 (more information available here: https://imotions.com/blog/eye-tracker-prices/), but ultimately getting one for under $400 will require building it yourself. Webcam versions might seem like an option, but they are simply not at the level to detect saccades, so you would miss this information (the accuracy of webcam-based methods just aren't at a good enough level yet). There are of course some guides that could help in building your own (e.g. Article How to Build a Low-Cost Eye-Tracking System
).Just wanted to add that info, and I hope The Eye Tribe tracker has served you well!
- Badges
- Science topic
- Similar topics
- Psychology