GEDAI Tutorials

This gallery is organized as a learning path so users can start quickly and then move to advanced workflows.

How to use this gallery

1. Start with the basics section to understand model behavior and core parameters.

2. Continue with practical pipelines for recommended and reusable denoising pipelines.

3. Use advanced if you want to dive into online processing and custom reference covariance workflows.

Basics: Understanding GEDAI Models

This section introduces the core GEDAI methods and when to use each one.

• How the standard Gedai model separates signal and noise subspaces.

• How MultibandGedai extends denoising to frequency-specific artifacts.

• How AdaptiveMultibandGedai adapts epoch duration across wavelet bands.

Understanding the GEDAI model

Understanding the GEDAI model

Understanding the multiband extension of GEDAI

Understanding the multiband extension of GEDAI

Understanding the adaptive extension of multiband GEDAI

Understanding the adaptive extension of multiband GEDAI

Practical Pipelines

This section contains end-to-end templates designed to be copied and adapted to your own data.

• 00_gedai_offline_pipeline.py: recommended offline denoising pipeline.

Recommended Offline EEG Denoising Pipeline

Recommended Offline EEG Denoising Pipeline

Advanced Workflows

This section covers specialized workflows for users who already know the basic GEDAI pipeline.

• 00_gedai_online.py: online denoising with chunked/epoch-wise processing.

• 10_gedai_forward.py: building a custom reference covariance from an MNE forward model.

Using GEDAI for online (real-time) denoising

Using GEDAI for online (real-time) denoising

GEDAI forward

GEDAI forward