Speech communication system and method, and robot apparatus

A speech communication apparatus enabling a conversation with a conversation partner, comprising:

• a speech recognition unit configured to recognize a speech content of the conversation partner;
• an audio output unit configured to output auditory communications;
• an image recognition unit configured to recognize a face of the conversation partner;
• a touch sensing unit configured to recognize a touch input by the conversation partner;
• a tracking control unit configured to determine whether or not to continue the conversation based on a recognition result from the image recognition unit or the touch sensing unit;
• and
• a network interface configured to communicate with an external network.

A speech communication method enabling a conversation with a conversation partner, said method comprising:

• generating a plurality of auditory communications according to a predetermined rule;
• recognizing, using a speech recognition unit in an apparatus, a speech content of the conversation partner;
• estimating intentions of the conversation partner from the speech content recognized by the recognizing of the speech recognition unit;
• dynamically selecting one of the plurality of auditory communications based on the estimation by the estimating;
• outputting the one of the plurality of auditory communications selected by the dynamically selecting;
• recognizing, using an image recognition unit in the apparatus, a face of the conversation partner;
• recognizing, using a touch sensing unit in the apparatus, a touch input by the conversation partner;
• and
• determining whether or not to continue the conversation based on a recognition result from recognizing by the image recognition unit or the touch sensing unit.

A non-transitory computer readable medium having stored thereon a program that when executed by a computing device causes the computing device to implement a speech communication method enabling a conversation with a conversation partner, said method comprising:

• generating a plurality of auditory communications according to a predetermined rule;
• recognizing, using a speech recognition unit in an apparatus, a speech content of the conversation partner;
• estimating intentions of the conversation partner from the speech content recognized by the recognizing by the speech recognition unit;
• dynamically selecting one of the plurality of auditory communications based on the estimation by the estimating;
• outputting the one of the plurality of auditory communications selected by the dynamically selecting;
• recognizing, using an image recognition unit in the apparatus, a face of the conversation partner;
• recognizing, using a touch sensing unit in the apparatus, a touch input by the conversation partner;
• and
• determining whether or not to continue the conversation based on a recognition result from recognizing by the image recognition unit or the touch sensing unit.