知搜-全球专利检索

  1. 登录
  2. 注册
发明授权
CN107039046B 一种基于特征融合的语音声效模式检测方法 失效 - 权利终止

一种基于特征融合的语音声效模式检测方法
摘要：
本发明公开了一种基于特征融合的语音声效模式检测方法。该方法包括如下步骤：接收语音信号；检测出所述语音信号中的元音，生成元音集合；提取元音集合中每一个元音的谱特征矢量序列；提取元音集合中每一个元音的频率特征矢量序列；根据元音集合中每一个元音的谱特征矢量序列和频率特征矢量序列生成该元音的声效特征矢量序列；将元音集合分别与多个候选声效模式进行匹配，生成每一个候选声效模式的匹配值；将匹配值最大的候选声效模式确定为语音信号的声效模式。
摘要（英）：

The invention discloses a voice sound effect mode detection method based on characteristic fusion. The method comprises steps that a voice signal is received; vowels of the voice signal are detected, and a vowel set is generated; a spectrum signature vector sequence of each vowel in the vowel set is extracted; a frequency characteristic vector sequence of each vowel in the vowel set is extracted; a sound effect characteristic vector sequence of the vowel is generated according to the spectrum signature vector sequence and the frequency characteristic vector sequence of each vowel in the vowel set; the vowel set and multiple candidate sound effect modes are respectively matched to generate matching values of each candidate sound effect mode; the candidate sound effect mode having the largest matching value is determined to be a sound effect mode of the voice signal.

公开/授权文献
0/0