摘要 |
Database stores reference voice data for beginner's, intermediate and advance levels. Text in lesson screen displayed on CRT is selected, reference voice data corresponding to this text is read out and model pronunciation is generated. User listens to this, and imitates pronunciation. Computer obtains voice data through the spectrum analysis of the user voice by voice recognition unit and determines user pronunciation level. Predetermined success mark is displayed on screen, if user pronunciation is so good that it is communicated exactly to collocutor. If determination result is bad, practice is repeated for the same text many times. This allows user to judge if his/her pronunciation is recognized by foreigner and improve foreign language pronunciation learning effect, by repeating this practice.
|