The Implementation and Evaluation of Individual Preference in Robot Facial Expression Based on Emotion Estimation...

Area of Science:

• Human-Robot Interaction
• Affective Computing
• Robotics

Background:

• Robot services are increasingly common, necessitating user acceptance for effective interaction.
• Previous attempts to synchronize robot facial expressions with human emotions showed varied user perceptions.
• Individual differences in user preferences significantly impact robot acceptance.

Purpose of the Study:

• To enhance robot acceptance by personalizing robot expressions based on user emotions.
• To investigate the impact of different robot expression strategies (synchronization, inverse synchronization, funny) on user perception.
• To develop a classification model for personalized robot expressions using biological signals and user feedback.

Main Methods:

• Estimating user emotion via biological signals.
• Implementing three robot expression conditions: synchronized, inversely synchronized, and funny.
• Collecting user feedback using "like/dislike" and the Semantic Differential scale.
• Utilizing logistic regression to build a personalized expression classification model.

Main Results:

• Robot expressions based on inverse synchronization during negative user emotions led to varied individual impressions.
• A personalized robot expression model, considering individual differences, resulted in the highest user acceptance.
• The proposed personalized method significantly improved user acceptance compared to fixed expression strategies.

Conclusions:

• Personalized robot expressions are crucial for optimizing user acceptance in human-robot interaction.
• Adapting robot expressions to individual emotional states and preferences enhances the effectiveness of robot services.
• Future robot designs should incorporate adaptive and personalized affective capabilities.