摘要 |
PROBLEM TO BE SOLVED: To extract an accurate command by properly recognizing inputted speech. SOLUTION: This command output device 1 is provided with a sound receiving part 3 receiving speech and converting the received speech into language, a command grammar dictionary part 4 storing vocabulary and grammar constituting commands, a verification dictionary part 5 storing features in regard to speaking, a comparing part checking the speech received at the sound receiving part 3 against the command grammar dictionary part 4 and the verification dictionary part 5, and a transmitting part 7 sending out the command to an equipment 2. The speech received at the sound receiving part 3 is checked against the command grammar dictionary part 4 and a matching degree with the command grammar dictionary part 4 is calculated. And the speech is checked against the verification dictionary part 5 and a matching degree with the verification dictionary part 5 is calculated. When a ratio of the calculated matching degrees with the command grammar dictionary part 4 and the verification dictionary part 5 is not less than a preset predetermined value, the inputted speech is extracted as a command and the extracted command is sent out to the equipment 2 by the transmitting part 7.
|