发明名称 Speech signal processing apparatus and method
摘要 A speech signal processing apparatus includes an amplitude and phase signal generation section that, based on an analyzing signal expressed by a complex signal generated from a speech signal applied with pitch marks every 1 pitch cycle, generates an amplitude signal and a phase signal on the time axis of the speech signal, a phase signal conversion section that converts the phase signal into a phase signal of a target pitch cycle width for each section of the 1 pitch cycle width based on the pitch marks, and a pitch conversion speech signal generation section that generates a speech signal in which pitch cycle is converted to the target pitch cycle based on an amplitude signal of the target pitch cycle width of a section corresponding to the section of the amplitude signal and based on a phase signal of the target pitch cycle width.
申请公布号 US9257131(B2) 申请公布日期 2016.02.09
申请号 US201314067446 申请日期 2013.10.30
申请人 FUJITSU LIMITED 发明人 Watanabe Kazuhiro
分类号 G10L19/00;G10L21/00;G10L21/013 主分类号 G10L19/00
代理机构 Staas & Halsey LLP 代理人 Staas & Halsey LLP
主权项 1. A speech signal processing apparatus comprising: a processor configured to generate, based on an analyzing signal expressed by a complex signal generated from a speech signal to which pitch marks are applied per 1 pitch cycle, an amplitude signal and a phase signal on a time axis of the speech signal; convert the generated phase signal into a phase signal of a target pitch cycle width per section of a 1 pitch cycle width based on the pitch marks; and generate a speech signal in which a pitch cycle is converted to the target pitch cycle based on an amplitude signal of the target pitch cycle width of a section corresponding to the section of the generated amplitude signal and the converted phase signal of the target pitch cycle width.
地址 Kawasaki JP