The Design of CNN Architectures for Optimal Six Basic Emotion Classification Using Multiple Physiological Signals.

Area of Science:

• Affective computing
• Biomedical engineering
• Machine learning for emotion recognition

Background:

• Accurate emotion recognition is crucial for human-computer interaction and mental health.
• Existing methods often struggle with individual variability in emotional responses.
• Multimodal physiological signals offer a promising avenue for enhanced emotion detection.

Purpose of the Study:

• To design an optimal emotion recognition method using multiple physiological signal parameters.
• To improve the accuracy of classifying individual emotional responses.
• To investigate the effectiveness of a deep learning approach for emotion classification.

Main Methods:

• Acquired respiration (RSP) and heart rate variability (HRV) signals from 53 participants across six basic emotions.
• Utilized two RSP parameters and five HRV parameters from specialized sensors.
• Developed and applied a novel convolutional neural network (CNN) deep learning model.
• Explored signal combinations and parameter influence on classification accuracy.

Main Results:

• Achieved high classification accuracy for individual emotions using the proposed CNN model.
• Demonstrated the effectiveness of combining multiple physiological signal parameters.
• Identified dominant factors influencing emotion classification accuracy.

Conclusions:

• The developed deep learning model effectively recognizes emotions using multimodal physiological signals.
• Combining RSP and HRV parameters enhances emotion classification accuracy.
• This method provides a foundation for future advancements in CNN-based emotion recognition.