公开(公告)号：US10770065B2

公开(公告)日：2020-09-08

申请号：US15842430

申请日：2017-12-14

Applicant: Samsung Electronics Co., Ltd.

Inventor： Sang-yoon Kim ,  Sung-soo Kim ,  Il-hwan Kim ,  Kyung-min Lee ,  Nam-hoon Kim ,  Jong-youb Ryu ,  Jae-won Lee

IPC: G10L15/197 ,  G10L15/22 ,  G10L15/30

Abstract: A speech recognition method and a speech recognition apparatus which pre-download a speech recognition model predicted to be used and use the speech recognition model in speech recognition is provided. The speech recognition method, performed by the speech recognition apparatus, includes determining a speech recognition model, based on user information downloading the speech recognition model, performing speech recognition, based on the speech recognition model, and outputting a result of performing the speech recognition.

公开(公告)号：US10171911B2

公开(公告)日：2019-01-01

申请号：US15531916

申请日：2015-11-27

Applicant: SAMSUNG ELECTRONICS CO., LTD.

Inventor： Yoon-jae Lee ,  Han-ki Kim ,  Sang-yoon Kim ,  Young-suk Song ,  Eun-mi Oh ,  Dong-hyun Lim ,  Ji-ho Chang ,  Jae-youn Cho

IPC: H04R5/04 ,  H04R5/02 ,  H04R3/12

Abstract: Provided is a method of processing an audio signal, the method performed by a device and including dividing the audio signal into a main signal and a background signal; obtaining position information about a main speaker and an auxiliary speaker; mixing the main signal and the background signal, based on the position information; and outputting the mixed main signal and the mixed background signal to the main speaker and the auxiliary speaker, respectively.

公开(公告)号：US09323982B2

公开(公告)日：2016-04-26

申请号：US14093888

申请日：2013-12-02

Applicant: SAMSUNG ELECTRONICS CO., LTD.

Inventor： Sang-yoon Kim ,  Ki-jun Jeong ,  Hee-seob Ryu ,  Mikiyas Teshome ,  Seung-kwon Park

IPC: G06K9/00 ,  G06F21/32 ,  H04N21/475 ,  H04N21/4223 ,  H04N21/4415

CPC classification number: G06K9/00288 ,  G06F21/32 ,  G06F2221/2149 ,  G06K9/00248 ,  H04N21/4223 ,  H04N21/4415 ,  H04N21/4751

Abstract: A method of certifying a user is provided. The method includes: generating a learning system; if a reference image is registered, analyzing the reference image by using the learning system to detect characteristic information; storing the detected characteristic information; photographing a user to acquire a user image; analyzing the user image by using the learning system to detect face characteristic information and additional information of the user; and comparing the face characteristic information and the additional information with stored characteristic information to certify the user. Therefore, a user certification is accurately performed.

Abstract translation: 提供了一种验证用户的方法。 该方法包括：生成学习系统; 如果注册参考图像，则通过使用学习系统来分析参考图像以检测特征信息; 存储检测到的特征信息; 拍摄用户以获取用户图像; 通过使用学习系统来分析用户图像来检测用户的面部特征信息和附加信息; 以及将所述面部特征信息和所述附加信息与存储的特征信息进行比较以证明所述用户。 因此，准确地执行用户认证。