Title: Time-domain blind separating method for in-line natural voice convolution mixing signal
Application Number: 200610030252 Application Date: 2006.08.22
Publication Number: 1909064 Publication Date: 2007.02.07
Approval Pub. Date: Granted Pub. Date:
International Classifi-cation: G10L15/28;G10L15/22;G10L15/00;H03H17/02
Applicant(s) Name: Fudan University Address:
Inventor(s) Name: Lu Xiaodan;Zhang Liming
Attorney & Agent: lu fei sheng zhifan
Abstract:
    The related unsupervised learning algorithm to on-line treat speech signal for speech interactive front-end process comprises: with sufficient respect to time-domain relevance and non-stationary of natural speech signal, constructing a group of adaptive feed-forward filter networks to separate speech signal from multiple sources. This invention enhances speech signal, reduces algorithm complexity for hardware implementation, and has wide application.
Time: 10