This project focuses on processing EEG signals from the DEAP dataset to recognize human emotions. The pipeline includes feature extraction, and classification using deep learning and machine learning models.
The DEAP dataset is a multimodal dataset for the analysis of human affective states, containing EEG and peripheral physiological signals recorded from 32 participants while watching 40 one-minute music videos.
- Signal segmentation and normalization
- Feature extraction (statistical, frequency-based)
- Emotion labeling based on arousal/valence scores
- Deep learning models for emotion classification
- Visualization of results
In this project, we focus on specific EEG frequency bands that are known to be associated with different mental states:
- Alpha (8–12 Hz): Associated with relaxed, calm, and restful states. Often seen when the eyes are closed.
- Beta (12–30 Hz): Linked to active thinking, focus, and concentration.
- Gamma (30–64 Hz): Related to high-level cognitive functioning, perception, and consciousness.
These bands are extracted from the EEG signals to serve as features for emotion classification.