公开(公告)号：US20160148615A1

公开(公告)日：2016-05-26

申请号：US14946021

申请日：2015-11-19

Applicant: Samsung Electronics Co., Ltd.

Inventor： Tae-Jin LEE ,  Sang-Hoon LEE ,  Subhojit CHAKLADAR

IPC: G10L15/22 ,  G06F1/32 ,  G06F3/16

CPC classification number: G10L15/22 ,  G06F1/3215 ,  G06F1/3231 ,  G06F1/324 ,  G06F1/3265 ,  G06F1/3293 ,  G06F3/167 ,  G10L2015/088 ,  G10L2015/223 ,  Y02D10/122 ,  Y02D10/126 ,  Y02D10/153 ,  Y02D10/173

Abstract: Disclosed are a method and electronic device for voice recognition. The voice recognition method includes recognizing, in a first processor using low power mode, a voice signal inputted through a microphone, entering an active state and performing voice recording in a second processor if the recognized voice signal is a previously set keyword, and performing voice recognition in the second processor if the end of a voice input is determined during the voice recording.

Abstract translation: 公开了一种用于语音识别的方法和电子设备。 语音识别方法包括：在第一处理器中使用低功率模式，通过麦克风输入的语音信号，进入活动状态，并且如果识别的语音信号是先前设定的关键字，则在第二处理器中执行语音记录，并且执行语音 在语音记录期间确定语音输入的结束时在第二处理器中的识别。