在做语音研究时,无论是语音信号处理、语音识别、语音合成还是语义理解,也无论采用的方法是波形拼接还是参数合成,一个标准的语音材料是必不可少的。国内一些从事语音研究的单位往往只做一些适合自己课题的语音材料,或者根本苦于找不到现成适用的语音素材, When doing phonetic research, whether it is speech signal processing, speech recognition, speech synthesis or semantic understanding, a standard speech material is indispensable whether the method used is waveform splicing or parametric synthesis. Some domestic units engaged in voice research tend to only do some suitable for their own subject of voice material, or simply can not find readily available voice material,