US Patent Issued to Canon on May 13 for "Training of neural networks in which deformation processing of training data is adjusted so that deformed character images are not too similar to character images of another class" (Japanese Inventor)

ALEXANDRIA, Va., June 12 -- United States Patent no. 12,300,010, issued on May 13, was assigned to Canon K.K. (Tokyo).

"Training of neural networks in which deformation processing of training data is adjusted so that deformed character images are not too similar to character images of another class" was invented by Keisui Okuma (Kanagawa, Japan).

According to the abstract* released by the U.S. Patent & Trademark Office: "In a scene where a pseudo character image is generated by performing deformation processing for a character image, a character image that impedes training is suppressed from being generated. Based on a condition relating to a parameter that is used for the deformation processing and associated with a first class, a parame...