US 12,364,382 B2
Method for controlling a home appliance
Michael Rupp, Holzheim (DE); Kai Paintner, Welden (DE); and Kuldeep Narayan Singh, Dillingen a.d. Donau (DE)
Assigned to BSH Hausgeraete GmbH, Munich (DE)
Appl. No. 17/639,658
Filed by BSH Hausgeraete GmbH, Munich (DE)
PCT Filed Aug. 25, 2020, PCT No. PCT/EP2020/073695
§ 371(c)(1), (2) Date Mar. 2, 2022,
PCT Pub. No. WO2021/043624, PCT Pub. Date Mar. 11, 2021.
Claims priority of application No. 19195575 (EP), filed on Sep. 5, 2019.
Prior Publication US 2022/0287540 A1, Sep. 15, 2022
Int. Cl. A47L 15/00 (2006.01); A47L 15/42 (2006.01); A47L 15/46 (2006.01); G06N 3/045 (2023.01); G06T 3/00 (2024.01); G06T 3/4046 (2024.01); G06T 3/4053 (2024.01); G06T 3/4076 (2024.01); G06T 7/00 (2017.01); G06T 7/10 (2017.01); H04L 12/28 (2006.01)
CPC A47L 15/0063 (2013.01) [A47L 15/4295 (2013.01); A47L 15/46 (2013.01); G06N 3/045 (2023.01); G06T 3/00 (2013.01); G06T 3/4046 (2013.01); G06T 3/4053 (2013.01); G06T 7/00 (2013.01); G06T 7/10 (2017.01); H04L 12/2816 (2013.01); A47L 2401/04 (2013.01); A47L 2401/34 (2013.01); A47L 2501/04 (2013.01); A47L 2501/06 (2013.01); A47L 2501/07 (2013.01); A47L 2501/20 (2013.01); A47L 2501/30 (2013.01); G06T 3/4076 (2013.01)] 13 Claims
OG exemplary drawing
 
1. A method for controlling a home appliance in dependence on a load to be handled by the home appliance, the method comprising:
taking a load image of the load with a camera disposed at the home appliance, wherein the camera provides the load image with a resolution of less than 20,000 pixels;
sending the load image from the home appliance to a server;
receiving the load image on the server;
processing the load image on the server by upscaling the load image to obtain an upscaled load image using a trained generative adversarial network, and then segmenting the load in the upscaled load image with a trained neural network to obtain a labelled segmented image indicating a segmentation of the load according to types of the load; and
based on the labelled segmented image, dynamically adapting a water pressure of a water spray of a spray arm in the home appliance according to one of the types of the load at which the water spray is presently targeted;
wherein weights and biases of the trained neural network have been fine-tuned using upscaled load images.