The task of social skills execution involves deciding what response actions should be generated by the robot, given the recognised current social state as described in the previous section. Such actions include both communicative actions (i.e., dialogue acts, such as greeting or asking a customer for their order), social actions (such as managing queueing), and non-communicative actions (typically, serving a drink); the system must also decide how communicative actions are realised, i.e., which combinations of modalities should be used (speech and/or gestures). This decision-making process should lead to robot behaviour that is both task-e ective and socially appropriate. An additional challenge is to make this decision-making robust to the generally incomplete and noisy observations that social state recognition is based on.
The task of social skills execution involves deciding what response actions should be generated by the robot, given the recognised current social state as described in the previous section. Such actions include both communicative actions (i.e., dialogue acts, such as greeting or asking a customer for their order), social actions (such as managing queueing), and non-communicative actions (typically, serving a drink); the system must also decide how communicative actions are realised, i.e., which combinations of modalities should be used (speech and/or gestures). This decision-making process should lead to robot behaviour that is both task-e ective and socially appropriate. An additional challenge is to make this decision-making robust to the generally incomplete and noisy observations that social state recognition is based on.
การแปล กรุณารอสักครู่..