US 12,249,122 B2
Holographic display calibration using machine learning
Manoj Sharma, Troy, MI (US); Thomas A. Seder, Fraser, MI (US); and Kai-Han Chang, Sterling Heights, MI (US)
Assigned to GM GLOBAL TECHNOLOGY OPERATIONS LLC, Detroit, MI (US)
Filed by GM Global Technology Operations LLC, Detroit, MI (US)
Filed on Aug. 3, 2022, as Appl. No. 17/817,061.
Prior Publication US 2024/0046619 A1, Feb. 8, 2024
Int. Cl. G06V 10/77 (2022.01); G02B 27/01 (2006.01); G06V 10/82 (2022.01)
CPC G06V 10/7715 (2022.01) [G02B 27/0103 (2013.01); G06V 10/82 (2022.01)] 20 Claims
OG exemplary drawing
 
1. A system comprising a computer including a processor and a memory, the memory including instructions such that the processor is programmed to:
receive, at a deep neural network, an image captured by an image capture device, wherein the image is projected onto a screen and includes distortion that is created by imperfections and surface disproportion along the screen;
generate, by the deep neural network, a predicted distortion map based on the image, wherein the predicted distorted map is estimated based on one or more predefined points along the screen, and wherein the processor applies the predicted distortion map to the image captured by the image capture device to create a substantially undistorted image; and
update at least one weight of the deep neural network in response to determining a loss associated with the predicted distortion map generated by the deep neural network has been minimized by a loss function.