ALEXANDRIA, Va., Aug. 20 -- United States Patent no. 12,394,430, issued on Aug. 19, was assigned to SAMSUNG ELECTRONICS Co. Ltd. (Suwon-si, South Korea).
"Method and apparatus with decoding in neural network for speech recognition" was invented by Jinwoo Park (Seoul, South Korea), Min-Joong Lee (Suwon-si, South Korea), Jihyun Lee (Suwon-si, South Korea) and Hoshik Lee (Seongnam-si, South Korea).
According to the abstract* released by the U.S. Patent & Trademark Office: "A decoding method includes receiving an input sequence corresponding to an input speech at a current time; and in a neural network (NN) for speech recognition, generating an encoded vector sequence by encoding the input sequence, determining reuse tokens from candidate bea...