| 摘要 |
PROBLEM TO BE SOLVED: To provide a speech input device in which accuracy of speech recognition can be more improved. SOLUTION: A gain controller 10 sets the gain of an output adjusting amplifier 22 for amplifying the output of an audio signal to a speaker 3 in such a manner that the magnitude of the level of the audio signal component, which is estimated by an expressive speech extraction section 7, in a digital signal outputted by an A/D converter 6 attains the value smaller than the minimum value of the magnitude of the level of the expressive speech signal component, which is estimated in the past by the expressive speech extraction section 7, in the digital signal output by the A/D converter 6. The input gain controller 10 also sets the gain of an input adjusting amplifier 5 for amplifying the signal inputted to the A/D converter 6 in such a manner that the ratio of the magnitude of the level of the expressive speech signal component, which is estimated by the expressive speech extraction section 7, in the digital signal outputted by the A/D converter 6 to the maximum level outputtable by the A/D converter 6 attains 1:2. COPYRIGHT: (C)2006,JPO&NCIPI |
