US Patent Issued to Xerox on Sept. 16 for "Using multiple trained models to reduce data labeling efforts" (California, New York Inventors)

ALEXANDRIA, Va., Sept. 17 -- United States Patent no. 12,417,222, issued on Sept. 16, was assigned to Xerox Corp. (Norwalk, Conn.).

"Using multiple trained models to reduce data labeling efforts" was invented by Matthew Shreve (Mountain View, Calif.), Francisco E. Torres (San Jose, Calif.), Raja Bala (Pittsford, N.Y.), Robert R. Price (Palo Alto, Calif.) and Pei Li (San Jose, Calif.).

According to the abstract* released by the U.S. Patent & Trademark Office: "A method of labeling training data includes inputting a plurality of unlabeled input data samples into each of a plurality of pre-trained neural networks and extracting a set of feature embeddings from multiple layer depths of each of the plurality of pre-trained neural networks. The...