US 11,887,617 B2
Electronic device for speech recognition and control method thereof
Ki Hoon Shin, Suwon-si (KR); Jonguk Yoo, Suwon-si (KR); and Sangmoon Lee, Seongnam-si (KR)
Assigned to SAMSUNG ELECTRONICS CO., LTD., Suwon-si (KR)
Appl. No. 17/260,684
Filed by SAMSUNG ELECTRONICS CO., LTD., Suwon-si (KR)
PCT Filed May 31, 2019, PCT No. PCT/KR2019/006548
§ 371(c)(1), (2) Date Jan. 15, 2021,
PCT Pub. No. WO2020/017759, PCT Pub. Date Jan. 23, 2020.
Claims priority of application No. 10-2018-0083144 (KR), filed on Jul. 17, 2018.
Prior Publication US 2021/0327449 A1, Oct. 21, 2021
Int. Cl. G10L 21/0216 (2013.01); G10L 21/0272 (2013.01); G10L 25/18 (2013.01); G10L 25/60 (2013.01)
CPC G10L 21/0216 (2013.01) [G10L 21/0272 (2013.01); G10L 25/18 (2013.01); G10L 25/60 (2013.01); G10L 2021/02166 (2013.01)] 9 Claims
OG exemplary drawing
 
1. A control method of an electronic device, the control method comprising:
receiving a plurality of audio signals output from a plurality of sources through a plurality of microphones;
analyzing the plurality of audio signals and obtaining information on directions in which the plurality of audio signals is received and information on input times of the plurality of audio signals;
determining signal sizes and input duration times of the plurality of audio signals based on the information on input times of the plurality of audio signals;
identifying at least one audio signal having a signal size exceeding a predetermined signal size and having an input duration time exceeding a predetermined time among the plurality of audio signals;
determining whether each of the directions in which the plurality of audio signals are input, has been changed more than a predetermined angle among the identified at least one audio signal;
determining at least one source outputting an audio signal of which the determined direction in which the audio signal is input has been changed more than the predetermined angle;
determining a source outputting an audio signal having shortest input duration time among the at least one source as a target source; and
processing an audio signal obtained from the determined target source,
wherein the method further comprising:
based on at least one audio signal being input continuously for a preset time, increasing counter of the at least one audio signal, the counter representing continuity of the at least one audio signal,
determining a source outputting an audio signal of which counter is lowest among the plurality of the audio signals as the source outputting the audio signal having the shortest input duration time.