SPEECH PROCESSING APPARATUS AND METHOD USING A PLURALITY OF MICROPHONES

A device for processing a speech according to a technical idea of the present disclosure may comprise: a plurality of microphones that receive a plurality of input signals from the outside; a spatial filtering circuit that generates a spatial filter signal corresponding to the plurality of input sig...

Full description

Saved in:
Bibliographic Details
Main Authors SHIN JONG WON, HYUN DONG IL, KIM SEONG WOO, CHEONG SE IN
Format Patent
LanguageEnglish
Korean
Published 19.10.2021
Subjects
Online AccessGet full text

Cover

Loading…
More Information
Summary:A device for processing a speech according to a technical idea of the present disclosure may comprise: a plurality of microphones that receive a plurality of input signals from the outside; a spatial filtering circuit that generates a spatial filter signal corresponding to the plurality of input signals through spatial filtering; a static noise estimation circuit that generates estimation static noise information by combining the directional static noise information indicating a level of a static noise signal received in a direction of interest and the diffusion static noise information indicating the level of static noise signals received from multiple directions, based on whether or not the plurality of input signals have directionality; and a speech estimation circuit that generates an estimation speech signal by filtering the spatial filter signal based on the estimation static noise information. Therefore, the present invention is capable of having an effect of estimating reliable static noise information. 본 개시의 기술적 사상에 따른 음성 처리 장치는, 외부로부터 복수의 입력 신호들을 수신하는 복수의 마이크로폰들, 공간 필터링을 통해 상기 복수의 입력 신호들에 대응하는 공간 필터 신호를 생성하는 공간 필터링 회로, 상기 복수의 입력 신호들이 방향성을 갖는지 여부를 기초로, 관심 방향에서 수신되는 잡음 신호의 레벨을 나타내는 방향성 잡음 정보 및 여러 방향에서 수신되는 잡음 신호들의 레벨을 나타내는 확산성 잡음 정보를 통합함으로써 추정 잡음 정보를 생성하는 잡음 추정 회로 및 상기 추정 잡음 정보를 기초로 상기 공간 필터 신호를 필터링함으로써 추정 음성 신호를 생성하는 음성 추정 회로를 포함할 수 있다.
Bibliography:Application Number: KR20200043602