Robot emotion model generation method, emotion model and interactive method

Abstract

The invention discloses a robot emotion model generation method, an emotion model and an interactive method. The interactive method comprises the steps of presetting multiple different emotion types according to control or interactive demands, and taking the emotion types as observable state values; obtaining external factors capable of influencing robot emotions, and setting the external factorsto be in a hidden state; through simulation experiments or real person experiments, obtaining an initial state probability matrix and a state transfer probability matrix corresponding to preset different character types; through a forward-backward algorithm, building hidden Markov models corresponding to the different character types; through the hiding state and by adopting an expectation maximization algorithm, solving the hidden Markov models to obtain probabilities of all observable states; and selecting the emotion type corresponding to the observable state with the highest probability asthe emotion of a robot, and obtaining a robot emotion model. According to the scheme, relatively independent emotions can be established for the robot, so that the authenticity and experience of robot and user emotion interaction are improved.

Description

technical field [0001] The present invention relates to the technical field related to robot emotion interaction, in particular to a method for generating a robot emotion model, an emotion model and an interaction method. Background technique [0002] The concept of Affective Computing was proposed by Professor Picard of MIT Media Lab in 1997. Affective computing is essentially a typical pattern recognition problem. Smart machines use a variety of sensors to obtain various data such as human expressions, postures, gestures, voice, intonation, blood pressure, heart rate, etc., and combine contextual information such as the environment, context, and situation at that time to recognize and understand human emotions. Affective computing is a highly comprehensive technical field, and its main research contents can be summarized into six aspects: 1. Research on emotional mechanism; 2. Acquisition of emotional signals; 3. Analysis, modeling and recognition of emotional signals; 4....

AI Technical Summary

Problems solved by technology

However, the current emotional robots pay attention to the user's emotions, and then choose different reply methods for emotional interaction according to the user's emotions, so that the interaction is completely based on the user as the absolute subject. There will be inconsistencies in the actual situation of their respective emotions, resulting in the current emotional robot being completely a comforting robot, and users cannot obtain a real emotional interact

Images