US 11,989,345 B1
Machine learning based forecasting of human gaze
Gautam Prasad, Los Angeles, CA (US); Dmitry Lagun, San Jose, CA (US); and Florian Schroff, Santa Monica, CA (US)
Assigned to Google LLC, Mountain View, CA (US)
Appl. No. 18/548,439
Filed by Google LLC, Mountain View, CA (US)
PCT Filed May 28, 2021, PCT No. PCT/US2021/034714
§ 371(c)(1), (2) Date Aug. 30, 2023,
PCT Pub. No. WO2022/250683, PCT Pub. Date Dec. 1, 2022.
Int. Cl. G06F 3/01 (2006.01); G06T 7/70 (2017.01)
CPC G06F 3/013 (2013.01) [G06T 7/70 (2017.01); G06T 2207/20081 (2013.01); G06T 2207/20084 (2013.01); G06T 2207/30201 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A computer-implemented method comprising:
determining, based on an image representing an eye of a user, a measured eye gaze position of the eye;
determining a first incremental change in the measured eye gaze position by processing the measured eye gaze position by a long short-term memory (LSTM) model that has been trained to generate incremental changes in eye gaze positions;
determining a first predicted eye gaze position of the eye at a first future time based on the measured eye gaze position and the first incremental change in the measured eye gaze position;
determining a second incremental change in the first predicted eye gaze position by processing the first predicted eye gaze position by the LSTM model;
determining a second predicted eye gaze position of the eye at a second future time subsequent to the first future time based on the first predicted eye gaze position and the second incremental change in the first predicted eye gaze position; and
outputting one or more of: the first predicted eye gaze position or the second predicted eye gaze position.