Limits...
Smaller is better: drift in gaze measurements due to pupil dynamics.

Drewes J, Zhu W, Hu Y, Hu X - PLoS ONE (2014)

Bottom Line: Recently, a significant impact of changes in pupil size on gaze position as measured by camera-based eye trackers has been reported.We observed a wide range of drift direction, mostly downward and nasal.We demonstrate two methods to partially compensate the pupil-based shift using separate calibrations in pupil-constricted and pupil-dilated conditions, and evaluate an improved method of compensation based on individual look-up-tables, achieving up to 74% of compensation.

View Article: PubMed Central - PubMed

Affiliation: Centre for Vision Research, York University, Toronto, Canada; Center for Mind/Brain Sciences, Trento University, Rovereto, Italy.

ABSTRACT
Camera-based eye trackers are the mainstay of eye movement research and countless practical applications of eye tracking. Recently, a significant impact of changes in pupil size on gaze position as measured by camera-based eye trackers has been reported. In an attempt to improve the understanding of the magnitude and population-wise distribution of the pupil-size dependent shift in reported gaze position, we present the first collection of binocular pupil drift measurements recorded from 39 subjects. The pupil-size dependent shift varied greatly between subjects (from 0.3 to 5.2 deg of deviation, mean 2.6 deg), but also between the eyes of individual subjects (0.1 to 3.0 deg difference, mean difference 1.0 deg). We observed a wide range of drift direction, mostly downward and nasal. We demonstrate two methods to partially compensate the pupil-based shift using separate calibrations in pupil-constricted and pupil-dilated conditions, and evaluate an improved method of compensation based on individual look-up-tables, achieving up to 74% of compensation.

Show MeSH
Average normalized pupil diameter vs. screen background brightness.Mean and s.e.m. across subjects.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4206464&req=5

pone-0111197-g002: Average normalized pupil diameter vs. screen background brightness.Mean and s.e.m. across subjects.

Mentions: Gaze calibration is never perfect. To determine the overall quality of the calibration solutions derived by the polynomial fitting, these solutions were applied to the fixation positions measured during the calibration phase. The mean deviation of the recorded fixations from their respective reference grid positions were then averaged first between the eyes of each subject, then across subjects. As the accuracy of camera-based eye trackers is usually best around the screen center (straight forward viewing), the repetitive accuracy was computed separately for the entire calibration grid, spanning 20×20 degrees (5×5 = 25 points), for the center quadrant, spanning 10×10 degrees (3×3 = 9 points), and for the outer region (the remainder after removal of the center quadrant, 16 points). In all three cases, the mean deviation was not constant and increased with decreasing background brightness and thus increasing pupil dilation (ANOVA on full field data, df = 6, F = 76.2, p<0.0001); the mean deviation was more than twice as large in the black background/dilated pupil condition compared to the white background, constricted pupil condition (0.83±0.09 and 0.39±0.05, degree±sem, full field data). Repetitive accuracy in the center quadrant was consistently better than in outer regions, although the difference was small, averaging 0.07degree (paired t-test, df = 6, p<0.001). The fixation accuracy results can be seen in Figure 1, the average pupil size for each background brightness can be seen in Figure 2. In general, the shapes of the graphs in Figures 1 and 2 correspond neatly to each other, indicating a strong correlation between repetitive accuracy and pupil size.


Smaller is better: drift in gaze measurements due to pupil dynamics.

Drewes J, Zhu W, Hu Y, Hu X - PLoS ONE (2014)

Average normalized pupil diameter vs. screen background brightness.Mean and s.e.m. across subjects.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4206464&req=5

pone-0111197-g002: Average normalized pupil diameter vs. screen background brightness.Mean and s.e.m. across subjects.
Mentions: Gaze calibration is never perfect. To determine the overall quality of the calibration solutions derived by the polynomial fitting, these solutions were applied to the fixation positions measured during the calibration phase. The mean deviation of the recorded fixations from their respective reference grid positions were then averaged first between the eyes of each subject, then across subjects. As the accuracy of camera-based eye trackers is usually best around the screen center (straight forward viewing), the repetitive accuracy was computed separately for the entire calibration grid, spanning 20×20 degrees (5×5 = 25 points), for the center quadrant, spanning 10×10 degrees (3×3 = 9 points), and for the outer region (the remainder after removal of the center quadrant, 16 points). In all three cases, the mean deviation was not constant and increased with decreasing background brightness and thus increasing pupil dilation (ANOVA on full field data, df = 6, F = 76.2, p<0.0001); the mean deviation was more than twice as large in the black background/dilated pupil condition compared to the white background, constricted pupil condition (0.83±0.09 and 0.39±0.05, degree±sem, full field data). Repetitive accuracy in the center quadrant was consistently better than in outer regions, although the difference was small, averaging 0.07degree (paired t-test, df = 6, p<0.001). The fixation accuracy results can be seen in Figure 1, the average pupil size for each background brightness can be seen in Figure 2. In general, the shapes of the graphs in Figures 1 and 2 correspond neatly to each other, indicating a strong correlation between repetitive accuracy and pupil size.

Bottom Line: Recently, a significant impact of changes in pupil size on gaze position as measured by camera-based eye trackers has been reported.We observed a wide range of drift direction, mostly downward and nasal.We demonstrate two methods to partially compensate the pupil-based shift using separate calibrations in pupil-constricted and pupil-dilated conditions, and evaluate an improved method of compensation based on individual look-up-tables, achieving up to 74% of compensation.

View Article: PubMed Central - PubMed

Affiliation: Centre for Vision Research, York University, Toronto, Canada; Center for Mind/Brain Sciences, Trento University, Rovereto, Italy.

ABSTRACT
Camera-based eye trackers are the mainstay of eye movement research and countless practical applications of eye tracking. Recently, a significant impact of changes in pupil size on gaze position as measured by camera-based eye trackers has been reported. In an attempt to improve the understanding of the magnitude and population-wise distribution of the pupil-size dependent shift in reported gaze position, we present the first collection of binocular pupil drift measurements recorded from 39 subjects. The pupil-size dependent shift varied greatly between subjects (from 0.3 to 5.2 deg of deviation, mean 2.6 deg), but also between the eyes of individual subjects (0.1 to 3.0 deg difference, mean difference 1.0 deg). We observed a wide range of drift direction, mostly downward and nasal. We demonstrate two methods to partially compensate the pupil-based shift using separate calibrations in pupil-constricted and pupil-dilated conditions, and evaluate an improved method of compensation based on individual look-up-tables, achieving up to 74% of compensation.

Show MeSH