Speech recognition system

A speech recognition system that controls a device on a basis of a recognition result from a user's uttered speech, the speech recognition system comprising:

• a speech recognizer that continuously recognizes speech uttered by at least one user;
• a speech-operation-intention determinator that makes a determination on whether or not the user has an intention of operating the device by uttering the speech based on at least one of:
• (1) a volume, pitch, and articulation of the speech uttered by the user, (2) a rate of the user's speech, and (3) a direction of a source of the user's speech;
• and
• a control mode alterer that controls the device by sending instructions to the device based on the recognized speech and the determination of the speech-operation-intention determinator, wherein
• the speech-operation-intention determinator determines a reliability indicating a degree of the intention of operating the device when the speech-operation-intention determinator determines that the user has the intention of operating the device,
• when the speech uttered by the user is failed to be recognized, the instructions sent to the device cause the device to notify the user of the failure in recognizing the speech,
• the notification is altered based on the determination of the speech-operation-intention determinator such that when the reliability of the intention of operating the device is low, a mode of notification is different than the mode of notification in a case in which the reliability is high,
• when the speech uttered by the user is successfully recognized and it is determined that the user has no intention of operating the device with the uttered speech, the instructions sent to the device cause the device to operate in such a manner that the user is less aware of or pays less attention to the operation of the device than in a case in which it is determined that the user has the intention of operating the device,
• the control mode alterer obtains a determination result that indicates whether or not recognition of the speech by the user is unsuccessful,
• the control mode alterer controls the device, when the speech uttered by the user is failed to be recognized to inform the user of a failure in recognizing the speech;
• the control mode alterer changes the instructions sent to the device, when the speech-operation-intention determinator determines that the user has no intention of operating the device such that a mode of notification of the failed speech recognition is provided in such a manner that the user is less aware of or pays less attention to the mode of notification than in the case in which the speech-operation-intention determinator determines that the user has the intention of operating the device to be controlled,
• when the speech uttered by the user is failed to be recognized, the instructions sent to the device cause the device to notify the user of the failure in recognizing the speech by sound such that when the reliability of the intention of operating the device is low, the volume of the sound is lower than the volume of the sound notification when the reliability is high.