US 12,468,388 B2
System and method for eye-gaze direction-based pre-training of neural networks
Ron M. Hecht, Raanana (IL); Omer Tsimhoni, Bloomfield Hills, MI (US); Dan Levi, Ganei Tikvah (IL); Shaul Oron, Rehovot (IL); Andrea Forgacs, Kfar Sava (IL); Ohad Rahamim, Netanya (IL); and Gershon Celniker, Netanya (IL)
Assigned to GM Global Technology Operations LLC, Detroit, MI (US)
Filed by GM GLOBAL TECHNOLOGY OPERATIONS LLC, Detroit, MI (US)
Filed on Oct. 18, 2023, as Appl. No. 18/489,338.
Claims priority of provisional application 63/419,508, filed on Oct. 26, 2022.
Prior Publication US 2024/0143074 A1, May 2, 2024
Int. Cl. G06F 3/01 (2006.01); G06T 7/50 (2017.01); G06V 10/74 (2022.01)
CPC G06F 3/013 (2013.01) [G06T 7/50 (2017.01); G06V 10/761 (2022.01); G06T 2207/20084 (2013.01); G06T 2207/20228 (2013.01); G06V 2201/07 (2022.01)] 20 Claims
OG exemplary drawing
 
1. A method of training a disparity estimation network, the method comprising:
obtaining an eye-gaze dataset including a first plurality of images with at least one gaze direction associated with each of the first plurality of images;
training a gaze prediction neural network based on the eye-gaze dataset to develop a model trained to provide a gaze prediction for an external image;
obtaining a depth database including a second plurality of images having depth information associated with each of the second plurality of images, wherein the first plurality of images matches the second plurality of images.; and
training a disparity estimation neural network for object detection based on an output from the gaze prediction neural network and an output from the depth database.