US 11,657,803 B1
Method for speech recognition by using feedback information
Hyungwoo Kim, Gyeonggi-do (KR); and Dongchan Shin, Seoul (KR)
Assigned to ActionPower Corp., Seoul (KR)
Filed by ActionPower Corp., Seoul (KR)
Filed on Nov. 2, 2022, as Appl. No. 17/979,709.
Int. Cl. G10L 15/16 (2006.01); G10L 15/22 (2006.01)
CPC G10L 15/16 (2013.01) [G10L 15/22 (2013.01); G10L 2015/225 (2013.01)] 12 Claims
OG exemplary drawing
 
1. A speech recognition method performed by a computing device, the speech recognition method comprising:
performing first speech recognition on voice information to obtain first text information on the voice information;
receiving feedback regarding the first text information; and
generating final text information for the voice information based on the received feedback,
wherein the first speech recognition includes real-time speech recognition, and is performed through a first neural network model that is a Real-Time Speech-To-Text (RT-STT) neural network model including a transducer structure,
wherein the feedback regarding the first text information includes at least one of: feedback on a correct text of the first text information and feedback on an incorrect text of the first text information,
wherein the generating of the final text information for the voice information includes:
generating the final text information by performing second speech recognition on the voice information when the received feedback is an incorrect feedback,
and wherein the second speech recognition is performed by a second neural network model that is a Speech-To-Text (STT) neural network model including an encoder and decoder structure.