US 11,868,521 B2
Method and device for determining gaze position of user, storage medium, and electronic apparatus
Li He, Beijing (CN)
Assigned to Beijing Xiaomi Mobile Software Co., Ltd., Beijing (CN)
Filed by Beijing Xiaomi Mobile Software Co., Ltd., Beijing (CN)
Filed on Apr. 2, 2021, as Appl. No. 17/221,427.
Claims priority of application No. 202010622072.5 (CN), filed on Jun. 30, 2020.
Prior Publication US 2021/0405742 A1, Dec. 30, 2021
Int. Cl. G06F 3/01 (2006.01); G06F 3/04886 (2022.01); G06F 18/214 (2023.01)
CPC G06F 3/013 (2013.01) [G06F 3/04886 (2013.01); G06F 18/214 (2023.01)] 16 Claims
OG exemplary drawing
 
1. A method for determining a gaze position of a user that is applied to a terminal with a display screen, and the method comprising:
obtaining a target distance from the display screen to a target user;
obtaining a user image of the target user, where the user image includes a global image, a head image, and an eye image, and the global image is an image of a target space in front of the display screen;
determining a first space where eyes of the target user are located from a plurality of preset subspaces within the target space based on the target distance and the global image; and
determining a gaze position of the target user on the display screen corresponding to the first space and the user image of the target user based on a preset correspondence among the subspaces, user images, and screen coordinates on the display screen,
wherein the determining a gaze position of the target user on the display screen further comprises:
determining a pre-trained hierarchical coordinate prediction model for the first space which has been trained, where each of the subspaces have a pre-trained hierarchical coordinate prediction model which has been trained based on user images, screen coordinates of the gaze position of the user on the display screen, and a plurality of preset hierarchies; and
inputting the user image of the target user into the hierarchical coordinate prediction model several times based on the number of hierarchies corresponding to the hierarchical coordinate prediction model to obtain the gaze position of the target user on the display screen under the corresponding hierarchy for each time;
wherein increasing sub-areas into which the display screen is divided increases the number of the hierarchy.