US 12,190,907 B2
Method and system for automatic back-channel generation in interactive agent system
Jin Yea Jang, Suwon-si (KR); Sa Im Shin, Seoul (KR); Min Young Jung, Hwaseong-si (KR); and San Kim, Seongnam-si (KR)
Assigned to KOREA ELECTRONICS TECHNOLOGY INSTITUTE, Seongnam-si (KR)
Filed by Korea Electronics Technology Institute, Seongnam-si (KR)
Filed on Aug. 23, 2022, as Appl. No. 17/894,061.
Claims priority of application No. 10-2021-0114490 (KR), filed on Aug. 30, 2021.
Prior Publication US 2023/0069655 A1, Mar. 2, 2023
Int. Cl. G10L 25/63 (2013.01); G10L 15/22 (2006.01); G10L 25/39 (2013.01)
CPC G10L 25/63 (2013.01) [G10L 15/22 (2013.01); G10L 25/39 (2013.01); G10L 2015/223 (2013.01)] 17 Claims
OG exemplary drawing
 
1. An automatic back-channel generation method comprising:
receiving an utterance of a user;
predicting a back-channel and predicting a utterance sentiment feature by analyzing the received utterance of the user in a back-channel prediction model;
generating the predicted back-channel; and
outputting the generated back-channel,
wherein the back-channel prediction model is an AI model that is trained to predict a back-channel to express from the utterance of the user, and
wherein a training of the back-channel prediction model to predict the back-channel is performed at the same as a training of the back-channel prediction model to predict the utterance sentiment feature is performed.