发明名称
摘要 PROBLEM TO BE SOLVED: To improve quality of reproduced speech, while suppressing the number of samples of excitation signals, in speech encoding and decoding by using a code book. SOLUTION: A noise code book 73 composed of a plurality of noise signals to which cutoff frequencies different from each other are allocated, and in which each noise signal does not include a frequency component of the cutoff frequency or less, that is allocated to the noise signal, is built in the speech encoding device and the speech decoding device. As the noise signal to which high cutoff frequency is allocated is generally selected as an excitation signal in a time zone when a low frequency component is not included in input speech, a characteristic of a noise like section among residual signals of the input speech is effectively expressed. COPYRIGHT: (C)2009,JPO&INPIT
申请公布号 JP4905262(B2) 申请公布日期 2012.03.28
申请号 JP20070152816 申请日期 2007.06.08
申请人 发明人
分类号 G10L19/08;G10L19/12 主分类号 G10L19/08
代理机构 代理人
主权项
地址