Voice processing method and device and device for voice processing

The embodiment of the invention provides a voice processing method and device and a device for voice processing. The method comprises the following steps: acquiring voice data to be processed; carrying out sound source position estimation on the voice data, detecting a first jump point in the voice...

Full description

Saved in:
Bibliographic Details
Main Authors YAO SHENGYU, PAN YIQIAN
Format Patent
LanguageChinese
English
Published 05.01.2021
Subjects
ACOUSTICS
MUSICAL INSTRUMENTS
PHYSICS
SPEECH ANALYSIS OR SYNTHESIS
SPEECH OR AUDIO CODING OR DECODING
SPEECH OR VOICE PROCESSING
SPEECH RECOGNITION
Online AccessGet full text

Cover

More Information
Summary:The embodiment of the invention provides a voice processing method and device and a device for voice processing. The method comprises the following steps: acquiring voice data to be processed; carrying out sound source position estimation on the voice data, detecting a first jump point in the voice data, wherein the first jump point is a time point indicating that the sound source position in thevoice data changes; and segmenting the voice data based on change information of speaker representation features in the voice data and the first jump point to obtain a segmentation result. According to the embodiment of the invention, the accuracy of speaker segmentation can be improved. 本发明实施例提供了一种语音处理方法、装置和用于语音处理的装置。其中的方法包括：获取待处理的语音数据；对所述语音数据进行声源位置估计，检测所述语音数据中的第一跳变点，所述第一跳变点为表示所述语音数据中声源位置发生变化的时间点；基于所述语音数据中说话者表征特征的变化信息以及所述第一跳变点对所述语音数据进行分割，得到分割结果。本发明实施例可以提高说话者分割的准确率。
Bibliography:Application Number: CN202011063543