We have recorded eye-tracking data with a participant wearing mascara, and D-Lab has wrongly identified her eyelashes as her pupil. How do we fix that without going frame-by-frame?

Run the automatic pupil detection by restricting the mask to a smaller space (in other words, try to mask out the eye lashes). Don’t check the “only no detections” button and let it process the entire data session by using the new mask. This will overwrite your original pupil data which contains the eyelashes mistakenly detected as pupils. It will also probably bring your pupil detection percentage down. But by using the “only no detections” option, you can change the mask and incrementally add more pupil data. Additionally you could also change the eye tracking algorithm in options (from Edge V2 to Dark Areas) and re-run automatic pupil detection for each mask setting – make sure that you check the “only no detections” option when you do this in order to not overwrite previous detected pupils.  Of course how much of this process can be done automatically also depends on how well the eye camera was positioned during the recording session.

Last update:
2015-05-19 16:18
Ergoneers Gmbh
Average rating:0 (0 Votes)

You cannot comment on this entry

Chuck Norris has counted to infinity. Twice.

Records in this category


Sticky FAQs