VOICE RECOGNITION SYSTEM
When the voice recognition system starts operating, the commands that the system is able to recognize are displayed on the multi display.
A shortcut menu is provided on the voice recognition command screen, enabling the user to operate the system with fewer voice commands.
The natural speech technology used for this voice recognition system can recognize a user's natural speech as voice recognition commands ("natural speech" refers to naturally spoken utterances that include several keywords).
Natural speech technology enhances the usability because the user does not have to memorize voice recognition commands exactly.
Natural speech technology uses the system's Automatic Speech Recognizer (ASR) and Natural Language Understanding (NLU) to recognize voice recognition commands.
The ASR extracts keywords from a user's natural speech using a phonemic sequence (the units of sound that make up a phrase or sentence).
The NLU analyzes these extracted keywords based on their probable meaning, interpreting the intent of the user's utterances.
In order to interpret the intent of a user's speech, the NLU unit records key words in a data table.
The data table is organized by task, and the data is used to interpret the probable meaning of the user.
Even if the user does not say the indicated voice command ("Go Home"), speech with a similar meaning can be recognized (for example: "Let's go home" and "Take me home").
Tech Tips
The natural speech technology used by the voice recognition system does not function in the following cases.
When the key words used are not in the data table.
When inputting address elements such as the city name in the address input screen.
When a phrase includes several intentions (Ex. "Go home... actually, no. I want to call Bob.").
When spoken instructions are for a level that cannot be recognized by the system. (Ex. When the user attempts to carry out a task from a different level while another task is still in progress, such as trying to make a phone call while the Destination screen is being displayed after giving the 'Destination' command.)
*: If the system cannot recognize all the utterances, it changes the screen to the voice recognition command screen that suits the recognized task.