Surveillance System and method of identifying Speaker thereof
One embodiment of the present invention discloses a method for identifying a speaker comprising: a step of receiving metadata of a speech feature, which is object identification data extracted from an acoustic feature of the target speech data, from a camera of a surveillance area; and a step of ide...
Saved in:
Main Author | |
---|---|
Format | Patent |
Language | English Korean |
Published |
10.05.2023
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Summary: | One embodiment of the present invention discloses a method for identifying a speaker comprising: a step of receiving metadata of a speech feature, which is object identification data extracted from an acoustic feature of the target speech data, from a camera of a surveillance area; and a step of identifying an object corresponding to the metadata, which is a maximum similarity as a result of comparing of the metadata of the target speech data and the metadata for each object pre-stored in the database, as a speaker who is the subject of the target speech data. Therefore, the present invention enables a highly accurate speaker identification.
본 발명의 일 실시예는, 감시영역의 카메라로부터 타겟 음성데이터의 음향특징으로부터 추출된 객체 식별데이터인 음성특징의 메타데이터를 수신하는 단계; 및 상기 타겟 음성데이터의 메타데이터와 데이터베이스에 기저장된 객체별 메타데이터의 비교 결과 유사도가 최대인 메타데이터에 대응하는 객체를 상기 타겟 음성데이터의 주체인 화자로 식별하는 단계;를 포함하는 화자 식별 방법을 개시한다. |
---|---|
Bibliography: | Application Number: KR20210149890 |