An Approach to Emotion Recognition Using Brain Rhythm Sequencing and Asymmetric Features

Area of Science:

• Neuroscience
• Signal Processing
• Bioinformatics

Background:

• Emotional regulation is crucial during self-isolation to prevent mood swings.
• Electroencephalography (EEG) signals offer a viable method for efficient emotion recognition.
• Brain rhythm sequencing, inspired by bioinformatics, has been previously used for seizure detection.

Purpose of the Study:

• To develop an efficient method for emotion recognition using EEG signals.
• To identify optimal asymmetric features from brain rhythm sequences for accurate emotion classification.
• To enable the development of portable emotion-aware devices for self-isolation scenarios.

Main Methods:

• EEG data was analyzed using a brain rhythm sequencing approach.
• Asymmetric features were extracted from sequences generated by different channel data using similarity measures.
• The optimal feature was identified through evaluation for emotion recognition.
• Classification was performed using a music emotion recognition experiment and the public DEAP dataset.

Main Results:

• An optimal asymmetric feature was identified, yielding remarkable accuracy in emotion recognition.
• Classification accuracies of approximately 80-85% were achieved using minimal channel data (one pair of symmetrical channels).
• Emotion recognition demonstrated strong individual characteristics, suggesting the need for subject-dependent approaches.

Conclusions:

• Brain rhythm sequencing with optimized feature extraction provides an efficient method for emotion recognition from EEG.
• The approach allows for high accuracy with fewer resources, making it suitable for portable devices.
• This novel method offers a promising pathway for future emotional applications, particularly during self-isolation.