Gaze and saccade based graphical manipulation

1. A system for presenting graphics on a display device, wherein the system comprises:

an eye tracking device configured to determine a gaze point of a user on a display device; and

a processing device configured to:

combine a first, gaze point based, quality map generated based on the gaze point of the user and a second, image content based, quality map generated based on one or more image content based factors independent of the gaze point of the user to generate a combined quality map, and

cause a rendered image to be displayed on the display device, wherein quality of each area of the rendered image displayed on the display device is determined based, at least in part, upon the combined quality map.

2. The system of claim 1 , wherein the combined quality map comprises a plurality of elements, each element corresponding to an area of the rendered image and having a quality value determining quality of a corresponding area of the rendered image.

3. The system of claim 1 , wherein a quality value of an element of the first quality map is determined based on a distance between an area of the rendered image that corresponds to the element and a location of the gaze point of the user.

4. The system of claim 3 , wherein the first quality map is adjusted dynamically based on changes to the location of the gaze point.

5. The system of claim 1 , wherein the second quality map comprises a frequency map generated based on content of at least one previously rendered image.

6. The system of claim 5 , wherein the frequency map is generated based on at least one of spatial changes of the content in the at least one previously rendered image or temporal changes of the content in two or more previously rendered images.

7. The system of claim 5 , wherein combining the first quality map and the second quality map comprises generating the combined quality map with quality values of the combined quality map higher or equal to respective quality values of the frequency map.

8. The system of claim 1 , wherein the second quality map is generated based on a distance of an area on the rendered image to a center of the rendered image.

9. The system of claim 1 , wherein combining the first quality map and the second quality map comprises:

inputting quality values of the first quality map and the second quality map to a machine learning model; and

receiving, from the machine learning model based on the inputting, combined quality values for the combined quality map, each of the quality values associated with an element of the combined quality map.

10. The system of claim 1 , wherein combining the first quality map and the second quality map comprises generating quality values of the combined quality map by computing a weighted sum of quality values of the first quality map and respective quality values of the second quality map.

11. The system of claim 1 , wherein the processing device is further configured to combining the first quality map and the second quality map with a third quality map generated based on one or more factors independent of the gaze point of the user.

12. The system of claim 11 , wherein the second quality map or the third quality map is generated based on at least one of:

a geometry of the display device;

information extracted from intermediate data of a depth of field shader;

information extracted from intermediate data of a motion blur shader from a previously rendered image;

explicit labeling by content creators; or

quality maps of previously rendered images.

13. The system of claim 1 , wherein the quality of an area of the rendered image comprises one of a sample rate of the area or a resolution of the area.

14. A method for presenting images on a display device, the method comprising:

combining, by a processing device, a first, gaze point based, quality map generated based on a gaze point of a user on a display device and a second, image content based, quality map generated based on one or more image content based factors independent of the gaze point of the user to generate a combined quality map; and

causing, by the processing device, a rendered image to be displayed on the display device, wherein quality of the rendered image displayed on the display device is determined based, at least in part, upon the combined quality map.

15. The method of claim 14 , wherein the combined quality map is generated by combining the first quality map and the second quality map based on a first weight of the first quality map and a second weight of the second quality map.

16. The method of claim 14 , wherein the combined quality map is generated by combining the first quality map and the second quality map based on weights of the first quality map, wherein the first quality map comprises a plurality of elements corresponding to a plurality of areas of the rendered image, and wherein each element is associated with one of the weights.

17. The method of claim 14 , wherein the combined quality map is generated by combining the first quality map and the second quality map based on dynamically updated weights of the first quality map.

18. A non-transitory machine readable medium having instructions thereon for presenting images on a display device, the instructions executable by one or more processors to perform operations comprising:

combining a first, gaze point based, quality map generated based on a gaze point of a user on a display device and a second, image content based, quality map generated based on one or more image content based factors independent of the gaze point of the user to generate a combined quality map; and

causing a rendered image to be displayed on the display device, wherein quality of the rendered image displayed on the display device is determined based, at least in part, upon the combined quality map.

19. The non-transitory machine readable medium of claim 18 , wherein the combined quality map comprises a plurality of elements, each element corresponding to an area of the rendered image and having a quality value determining the quality of a corresponding area of the rendered image.

20. The non-transitory machine readable medium of claim 18 , wherein a quality value of an element of the first quality map is determined based on a distance between an area of the rendered image that corresponds to the element and a location of the gaze point of the user.