| 摘要 |
PROBLEM TO BE SOLVED: To provide a voice learning device first deciding a teacher pronouncing the nearest to oneself pronounce by a learner and thereafter, learning while modeling the teacher at the time of learning while comparing a recorded learner' s voice with a beforehand recorded teacher's voice. SOLUTION: A learner voice storage part 2 preserves temporarily the inputted learner voice. A teacher voice storage part 3 preserves beforehand the voices respectively pronounced by plural foreigner teachers A-J with voice characteristics different from each other related to the prescribed number of respective English sentences. A teacher text storage part 4 preserves beforehand the prescribed number of these sentences as the text data. A characteristic extraction part 6 extracts respectively the characteristics of the voices from the voices preserved in the learner voice storage part 2 and one teacher voice selected from the voices of the teachers A-J preserved in the teacher voice storage part 3. A characteristic comparison part 7 compares the characteristics of the extracted voices. A teacher decision part 8 compares the learner voice characteristic with the characteristics of the teacher voices in the characteristic comparison part 7, and decides one teacher that the characteristics of both voices are the nearest.
|
