Gaze detection offset for gaze tracking models
Examples are disclosed herein that relate to gaze tracking. One example provides a computing device including an eye-tracking system including an image sensor, a logic device, and a storage device comprising instructions executable by the logic device to track an eye gaze direction by acquiring an image of the eye via the eye-tracking system, and determining a determined location of a center of a lens of the eye from the image of the eye. The instructions are further executable to adjust the determined location of the center of the lens on a sub-pixel scale by applying a predetermined sub-pixel offset to the determined location of the center of the lens to produce an adjusted location of the center of the lens, to determine a gaze direction from the adjusted location of the center of the lens, and perform an action on a computing device based on the gaze direction.
1. A computing device, comprising:
an eye-tracking system comprising an image sensor;
a logic device; and
a storage device comprising instructions executable by the logic device to tracking an eye gaze direction by
acquiring an image of the eye via the eye-tracking system,
determining a determined location of a center of a lens of the eye from the image of the eye, the location being determined on a pixel scale,
after determining the determined location of the center of the lens on a pixel scale, adjusting the determined location of the center of the lens of the eye on a sub-pixel scale by applying a predetermined sub-pixel offset to the determined location of the center of the lens of the eye to produce an adjusted location of the center of the lens, the predetermined sub-pixel offset being based on a machine-trained model for determining a most probable sub-pixel offset;
determining a gaze direction from the adjusted location of the center of the lens; and
performing an action on a computing device based on the gaze direction.
2. The computing device of claim 1 , wherein the image sensor comprises a visible light camera, and wherein the instructions are executable to determine the determined location of the center of the lens by determining a boundary of an iris of the eye and determining a center of the iris based upon the boundary of the iris.
3. The computing device of claim 2 , wherein the instructions are further executable to apply the sub-pixel offset to the determined center of the iris.
4. The computing device of claim 1 , wherein the image sensor comprises an infrared camera, and wherein the instructions are executable to determine the determined location of the center of the lens of the eye by determining a boundary of a pupil of the eye and determining a center of the pupil based upon the boundary of the pupil.
5. The computing device of claim 4 , wherein the instructions are further executable to apply the sub-pixel offset to the determined center of the pupil.
6. The computing device of claim 1 , the instructions further executable to receive feedback regarding error between gaze direction and an intended interaction location on a user interface, and adjust the predetermined sub-pixel offset based upon the feedback.
7. The computing device of claim 6 , wherein the feedback comprises user input indicating an incorrect gaze determination.
8. On a computing device, a method of determining a gaze direction, the method comprising:
acquiring an image of the eye via the eye-tracking system,
determining a determined location of a center of a lens of the eye from the image of the eye, the determined location being determined on a pixel scale,
after determining the determined location of the center of the lens on a pixel scale, adjusting the determined location of the center of the lens of the eye on a sub-pixel scale by applying a predetermined sub-pixel offset to the determined location of the center of the lens of the eye to produce an adjusted location of the center of the lens, the predetermined sub-pixel offset being based on a machine-trained model for determining a most probable sub-pixel offset;
determining a gaze direction from the adjusted location of the center of the lens; and
performing an action based on the gaze direction.
9. The method of claim 8 , wherein determining the determined location of the center of the lens comprises determining a boundary of an iris of the eye and determining a center of the iris based upon the boundary of the iris.
10. The method of claim 9 , wherein adjusting the determined location of the center of the lens comprises applying the predetermined sub-pixel offset to the center of the iris.
11. The method of claim 8 , wherein determining the determined location of the center of the lens of the eye comprises determining a boundary of a pupil of the eye and determining a center of the pupil based upon the boundary of the pupil.
12. The method of claim 11 , wherein adjusting the determined location of the center of the lens comprises applying the predetermined sub-pixel offset to the center of the pupil.
13. The method of claim 8 , further comprising receiving feedback regarding error between gaze direction and an intended interaction location on a user interface, and adjusting the predetermined sub-pixel offset based upon the feedback.
14. The method of claim 13 , wherein the feedback comprises user input indicating an incorrect gaze determination.