ALEXANDRIA, Va., Aug. 12 -- United States Patent no. 12,387,718, issued on Aug. 12, was assigned to Amazon Technologies Inc. (Seattle).

"Removing bias from automatic speech recognition models using internal language model estimates" was invented by Nilaksh Das (Seattle), Monica Lakshmi Sunkara (San Jose, Calif.), Sravan Babu Bodapati (Fremont, Calif.), Jinglun Cai (Seattle), Devang Kulshreshtha (Montreal), Jeffrey John Farris (Crystal Lake, Ill.), Nicholas G Aldridge (Seattle), Srikanth Ronanki (San Jose, Calif.) and Katrin Kirchhoff (Seattle).

According to the abstract* released by the U.S. Patent & Trademark Office: "Bias may be removed from automatic speech recognition model predictions using internal language model estimates. Audio da...