Essentia: Boosting Artifact Removal from EEG through Semantic Guidance Utilizing Diffusion Model (ICASSP2025)

This is the Official PyTorch implementation of our ICASSP2025 paper "Essentia: Boosting Artifact Removal from EEG through Semantic Guidance Utilizing Diffusion Model".

Quick Start

Install

# build with python3.10
conda create --name Essentia python=3.10
conda activate Essentiaenv 
pip install -r requirments.txt

Usage

Before usage

1. We provide a data loading API in our code, which can be customized to suit the characteristics of specific datasets.
2. The detailed loading functions are available in Code/func.py and Code/TrainContrastive.py.

Run Model

1. For running the Essentia, you should use the command python Code/TrainContrastive.py