Fine-Tuning Strategies for Continual Online EEG Motor Imagery Decoding: Insights from a Large-Scale Longitudinal Study

This is the official repository for the paper Fine-Tuning Strategies for Continual Online EEG Motor Imagery Decoding: Insights from a Large-Scale Longitudinal Study .

Installation

Run pip install . to install the eeg-continual package.

Usage

Reproducing the results is a 3-step process.

To enhance reproducability, we have included checkpoints for the first subject.

1. Cross-subject source training

Run source_training.py --config basenet.yaml to run the source training.

Note: You may want to change the configuration file before to customize training (number of subjects, number of epochs, model logging, ..).

2. Subject-specific fine-tuning

Run finetuning.py --config finetuning.yaml to run the subject-specific fine-tuning.

Note: You need to have checkpoints in the ckpt folder before starting fine-tuning. Also: check the configuration file

3. Causal evaluation with OTTA

Run causal_evaluation.py --config causal_eval.yaml to run the final evaluation.

Note: You need to have checkpoints in the ckpt folder before starting the evaluation. Also: check the configuration file

If there are any questions, feel free to contact me.

Citation

If you find this repository useful, please cite our work:

@article{wimpff2025fine,
  title={Fine-Tuning Strategies for Continual Online EEG Motor Imagery Decoding: Insights from a Large-Scale Longitudinal Study},
  author={Wimpff, Martin and Aristimunha, Bruno and Chevallier, Sylvain and Yang, Bin},
  journal={arXiv preprint arXiv:2502.06828},
  year={2025}
}