| 摘要 |
<p>A database (16) stores reference voice data for beginner's, intermediate and advance levels for each text of a foreign language textbook, and when any one of texts in the lesson screen displayed on a CRT (26) is selected, the reference voice data corresponding to this text is read out and the model pronunciation is output from a voice synthesis unit (34). The user listens to this, and imitates the pronunciation facing a microphone (20). The computer obtains the voice data through the spectrum analysis of the user voice input from the microphone (20) by the voice recognition unit (22) and determines the user pronunciation level by comparing the same with the reference voice data of the database (16). A predetermined success mark is displayed on the screen, if the user pronunciation is so good that it is communicated exactly to the collocutor, and recognized by the collocutor, and it proceeds to the practice of the following text. If the determination result is bad, the practice is repeated for the same text many times. This allows the user to judge if his/her pronunciation is recognized by a foreigner and improve the foreign language pronunciation learning effect, by repeating this practice. <IMAGE></p> |
