-
Notifications
You must be signed in to change notification settings - Fork 4
Classify Fixations by Dispersion Threshold
classify_idt R Documentation
Classify each gaze point based on the dispersion of gaze points.
classify_idt(data, dispersion_threshold = 1.6, time_window = 250)
data Data frame of the eye tracking data we want to process
dispersion_threshold Dispersion threshold in degrees
time_window Time window in ms (minimum fixation duration)
Input data frame columns
gazeHasValue
: Boolean if there is valid gaze data
eyeDataRelativeTimestamp
: Timestamp of the data
gazeorigin_x
: X coordinates of the gaze origin
gazeorigin_y
: Y coordinates of the gaze origin
gazeorigin_z
: Z coordinates of the gaze origin
gazePoint_x
: X coordinates of the gaze point
gazePoint_y
: Y coordinates of the gaze point
gazePoint_z
: Z coordinates of the gaze point
Additional columns in output
eventIndex
: Index of the event in which the gaze point lies
eventType
: Type of the current event
eventDuration
: Duration of the current event
fixation_x
: X coordinate of the current fixation (if eventType is a fixation)
fixation_y
: Y coordinate of the current fixation (if eventType is a fixation)
fixation_z
: Z coordinate of the current fixation (if eventType is a fixation)
This is an implementation of the fixation classification by dispersion threshold as described by Llanes-Jurado et al. (2020, doi:10.3390/s20174956). It uses the distance-dispersion threshold to identify fixations with the specified threshold and minimum duration.
According to Llanes-Jurado et al. the acceptable parameter ranges for a HTC Vive Pro Eye HMD is between 1-1.6 degrees and 0.25-0.4 seconds with 1 degree and 0.25 seconds being the optimum. Due to the different precision of the HoloLens 2 eye tracker these thresholds most likely can't be transferred without adaption.
The input data frame with the classification columns added