Design and implementation of an affective computing for recognition and generation of behaviors in a robot

In this paper adapting a robot to a social approach is presented in the selection of behaviors for interaction in real environments, which represents an emotionally charged, causing attention to focus on the most relevant aspects of the surroundings for the realization of the software system; which is part of a branch called affective computing, where there is a classification for emotion exhibited by a system. In this classification, the machine is included within the application for displaying and perceiving simulated emotions. Determining the emotion is by using templates based on probability theory of Markov. In this project the user interaction function is modified for vary with respect to the emotional state of the agent, which is determined according to the environment in which it is. Based in the Kinect sensor, emotional states according to body language of people and positions is detected. The system recognizes that the compatibility of the emotional state has been entered against it by users or persons with whom the project is tested. After recognizing the emotional state; a robot mimics the movements of the human pretending to have such emotions.