US 12,032,929 B2
System and method for cross domain generalization for industrial artificial intelligence applications class
Aditya Srivastava, Delhi (IN); Sanjay Shekhawat, Nagaur (IN); Rushil Gupta, Chandigarh (IN); Sachin Kumar, Hesse (DE); Kamal Galrani, Hessen (DE); Rahul Prajapat, Frankfurt am Main (DE); Naga Sai Pranay Modukuru, Stuttgart (DE); Rishabh Agrahari, Mumbai (IN); Nihal Rajan Barde, Sindhudurga (IN); Arnab Kumar Mondal, New Delhi (IN); and Prathosh A.P, Mysore (IN)
Assigned to TVARIT GMBH, (DE)
Filed by TVARIT GMBH, Frankfurt am Main (DE)
Filed on Dec. 8, 2022, as Appl. No. 18/063,106.
Claims priority of application No. 202141058407 (IN), filed on Dec. 15, 2021.
Prior Publication US 2023/0185540 A1, Jun. 15, 2023
Int. Cl. G06N 20/00 (2019.01); G06F 3/0484 (2022.01); G06F 8/30 (2018.01); G06F 9/445 (2018.01); G06F 9/455 (2018.01); G06N 3/0499 (2023.01); G06N 3/08 (2023.01); G06N 5/02 (2023.01); G06N 5/04 (2023.01)
CPC G06F 8/311 (2013.01) [G06N 3/0499 (2023.01); G06N 3/08 (2013.01)] 26 Claims
OG exemplary drawing
 
1. A cross domain generalization system for industrial artificial intelligence (AI) applications, the cross domain generalization system comprising:
a hardware processor; and
a memory coupled to the hardware processor, wherein the memory comprises a set of program instructions in the form of a plurality of subsystems, configured to be executed by the hardware processor, wherein the plurality of subsystems comprises:
a target encoder subsystem configured to
obtain target data from a target machine product, wherein the target data is high dimensional original multi-channel time series data; and
generate lower dimensional data for the obtained target data using a target artificial intelligence (AI) model, wherein the generated lower dimensional data are corresponding to a plurality of target embeddings data, and wherein the plurality of target embeddings data comprise compressed representation for the original multi-channel time series data of the target machine product;
apply the plurality of target embeddings data into a source classifier AI model; and
a source classifier subsystem configured to predict a quality of the target machine product by generating a plurality of class labels for each of the plurality of target embeddings data based on a result of the source classifier AI model.