| 摘要 |
PURPOSE:To reduce the capacity of the memory used for the title device by commonly storing reference syllable patterns for comparison and basic syllable patterns as shared syllable patterns. CONSTITUTION:A voice recognizing section 1 recognizes inputted voices by successively synthesizing syllable patterns which are used as references for comparison and comparing the synthesized patterns with the inputted voices and an output voice synthesizing section 2 synthesizes voices corresponding to a sound outputting instruction by using the basic syllable patterns and outputs the synthesized voices. A memory 3 which stores voice patterns commonly stores the syllable patterns used as references for comparison and the basic syllable patterns as shared syllable patterns. When the syllable patterns to be used as references for comparison and basic syllable patterns used when voices are outputted are shared in such way, the required capacity of the memory can be reduced. |
