摘要 |
PROBLEM TO BE SOLVED: To provide an agent device for performing the smooth communication between a driver and an agent. SOLUTION: In a touch-panel-type display, a voice recognition button for starting voice recognition and at the same time, displaying an agent on the display is displayed. The CPU of the agent device displays the agent (S27) when no agents are being displayed on the display at the time the voice recognition button is pressed (S21), operates the voice recognition part, determines the action of the agent based on output from each kind of sensor or the like including the result of voice recognition, and allows the agent to perform the action (S33 and S35). When the agent is already displayed, the voice recognition part is activated, the action of the agent is determined based on output from each kind of sensor or the like including the result of voice recognition, and the action is performed by the agent (S33 and S35). |