US 12,353,912 B2
Task processing method, task processing device, and computer program product
Tianxiang Chen, Shanghai (CN); Yuhong Nie, Shanghai (CN); Sanping Li, Beijing (CN); Anzhou Hou, Shanghai (CN); and Zhen Jia, Shanghai (CN)
Assigned to Dell Products L.P., Round Rock, TX (US)
Filed by Dell Products L.P., Round Rock, TX (US)
Filed on May 27, 2022, as Appl. No. 17/826,670.
Claims priority of application No. 202210432110.X (CN), filed on Apr. 22, 2022.
Prior Publication US 2023/0342193 A1, Oct. 26, 2023
Int. Cl. G06F 9/48 (2006.01); G06N 3/045 (2023.01); G06N 3/08 (2023.01)
CPC G06F 9/4881 (2013.01) [G06N 3/045 (2023.01); G06N 3/08 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A task processing method, comprising:
deploying a first neural network at a first computing node of an edge portion of a processor-based machine learning system implemented in accordance with an edge-cloud architecture;
deploying a second neural network at a second computing node of a cloud portion of the processor-based machine learning system implemented in accordance with the edge-cloud architecture;
instantiating a feedback loop between the edge portion of the processor-based machine learning system and the cloud portion of the processor-based machine learning system, utilizing at least one network coupled between the edge and cloud portions;
receiving a task to be processed by the first neural network at the first computing node;
determining that the first neural network failed to process the task; and
sending the task to the second computing node via the feedback loop for processing by the second neural network at the second computing node, wherein the first neural network is of a first type having a first accuracy level, and the second neural network is of a second type different than the first type, the second neural network having a second accuracy level higher than the first accuracy level, and further wherein the first neural network is at least in part converted from the second neural network.