Advancing Emotionally Aware Child-Robot Interaction with Biophysical Data and Insight-Driven Affective Computing

Area of Science:

• Affective Computing
• Human-Robot Interaction
• Artificial Intelligence

Background:

• Emotionally aware machines are crucial for advancing human-robot interaction (HRI).
• Integrating neuroscience, psychology, and AI can create adaptive, emotion-aware systems.
• Child-robot interaction (CRI) requires systems that dynamically recognize and respond to emotional states.

Purpose of the Study:

• To investigate the integration of affective computing techniques using biophysical data for emotionally aware machines.
• To enhance child-robot interaction (CRI) through real-time emotion recognition and response.
• To develop adaptive systems capable of dynamically recognizing and responding to human emotional states.

Main Methods:

• A real-world CRI pilot study using the NAO robot.
• Facial expression analysis and speech emotion recognition for real-time emotion detection.
• Combining handcrafted and deep learning features for facial expressions; acoustic features for speech emotions.

Main Results:

• Real-time CRI emotion recognition achieved 85% accuracy for facial expressions and 83% for speech emotions.
• Offline evaluation using Dynamic Bayesian Mixture Model (DBMM) reached 92% accuracy for facial expressions and 98% for speech emotions.
• Robots successfully transitioned children's negative emotions to neutral or positive states, aiding emotional regulation.

Conclusions:

• Affective computing and biophysical data integration advance emotionally aware machines and CRI.
• Emotion-aware robots show potential in therapeutic and educational interventions for children.
• This research establishes a foundation for personalized, empathetic human-machine interactions.