Real-World EEG Emotion Recognition (SEED Pilot)

This small, focused repository explores real-world EEG emotion recognition in a compact setting.

Highlights:

• A streamlined notebook-driven pipeline (seed-paper-14-channel-kernel-7.ipynb) targeting a reduced-channel configuration (e.g., 14-channel setups inspired by SEED).
• Experiments with 3D-CNN kernels and temporal windows designed to keep the model light enough for realistic deployment while preserving performance.
• Serves as a sandbox for prototyping:
  • Different channel subsets
  • Kernel sizes
  • Data splits and augmentation strategies

While smaller than the main DEAP repository, this project is useful to show how the research ideas scale down to practical, lower-channel configurations, which is important for consumer-grade EEG devices.