eegemolib 0.1.1.post1

Unified Evaluation Framework for EEG-Based Emotion Recognition Algorithms

EEGEmoLib

EEGEmoLib is a plug-and-play EEG toolkit for EEG-based emotion analysis. It streamlines EEG emotion recognition workflows by combining dataset management, feature engineering, model training, and visualization utilities in one practical framework.

Installation

pip install eegemolib

Core Modules

• Datasets: configurable dataset loading through a unified YAML-based workflow
• Feature extraction: 17 handcrafted EEG features, including coherence
• Feature selection: 11 methods across filter, wrapper, and embedding strategies
• Recognition models: 15 representative EEG emotion recognition models
• Visualization: tools for inspecting EEG signals, feature distributions, and model behavior

Quick Start

From the repository root:

python -m eegemolib.engine.protocol --cfg src/eegemolib/cfg/experiments/seed_protocols/cross_session/deepconvnet_psd.yaml

Before running, update dataset.data_root in the experiment config to point to your local dataset directory.

If you want to run directly from source without installing the package, use:

PYTHONPATH=src python -m eegemolib.engine.protocol --cfg src/eegemolib/cfg/experiments/seed_protocols/cross_session/deepconvnet_psd.yaml

Resources

• Homepage: https://eegemolib.github.io
• Documentation: https://eegemolib-docs.readthedocs.io/en/latest/index.html#
• PyPI: https://pypi.org/project/eegemolib/
• GitHub: https://github.com/Schmerle-Cki/EEGEmoLib/tree/main

The homepage provides a concise visual overview of EEGEmoLib, while the documentation site contains detailed module references.