Speech enhancement method and apparatus

By deeply integrating voiceprint recognition technology with speech enhancement technology, and using cascaded or joint learning processing, the target user's speech is accurately labeled and targeted to be enhanced. This solves the problems of poor target speech extraction and insufficient voiceprint fidelity in existing technologies, and achieves high-definition and robust enhancement in complex scenarios.

CN122417062APending Publication Date: 2026-07-17HANGZHOU ROBAM APPLIANCES CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
HANGZHOU ROBAM APPLIANCES CO LTD
Filing Date
2026-05-15
Publication Date
2026-07-17

Smart Images

  • Figure CN122417062A_ABST
    Figure CN122417062A_ABST
Patent Text Reader

Abstract

本发明提供了一种语音增强方法和装置，其中，预先注册目标用户的声纹模板，该方法包括：获取混合语音信号，基于目标用户的声纹模板将混合语音信号的各个语音帧标记为表征目标用户语音的目标语音帧和表征干扰信号的干扰帧；采用级联式处理或联合学习式处理的方式，基于标记结果对混合语音信号进行定向增强，输出增强后的目标用户语音。该方式中，可以将声纹识别技术与语音增强技术深度融合，借助声纹特征的唯一性实现对目标用户语音的精准定位，引导增强算法定向优化目标语音信号、高效抑制各类干扰信号；同时，在增强过程中保障目标用户原始声纹特征的完整性，提升低信噪比场景下的技术鲁棒性。
Need to check novelty before this filing date? Find Prior Art