How purohith08/bci-iv-year-project works
This repository combines motor-imagery EEG collection, offline classifier training, and a real-time maze-game demo. The collection app presents trials and records subject sessions; separate training scripts process BCI Competition GDF data and build distinct classifiers; the game simulates EEG, classifies movement, and applies predictions to maze navigation while recording performance. The excerpts establish the game’s simulation and buffer, and training stages, but do not show the collection save path or a connection from trained artifacts to the game. Those links are therefore omitted. There is no substantive README overview.
Main entry points: bci_maze_game.py Maze navigation · DATA BCI GDF dataset · mi_eeg_system.py Trial collection app · FBCSP+MLP.py FBCSP + MLP
Components 14
EEG Collection
- Trial collection app
Data-collection/mi_eeg_system.py
Offline Training
- FBCSP + MLP
Signal processing/FBCSP+MLP.py - FBCSP + EEGNet + SVM
Signal processing/FBCSP+EEGNET&SVM.py - Hybrid MLP + XGBoost
Signal processing/train.py
Maze Game
- EEG simulation and buffer
Game-Test/bci_maze_game.py - Signal processing and prediction
Game-Test/bci_maze_game.py - Maze navigation
Game-Test/bci_maze_game.py - Trial metrics
Game-Test/bci_maze_game.py
EEG Data & Results
- Session recordings
Data-collection/EEG_Dataset - BCI GDF dataset
DATA - Saved model artifacts
models - Game session logs
Game-Test/bci_results
Other components
- Study participant
- Game participant
Connections 10
- Study participant → Trial collection app: completes trials
Data-collection/mi_eeg_system.py - Trial collection app → Session recordings: records sessions
- FBCSP + MLP → BCI GDF dataset: loads GDF epochs
Signal processing/FBCSP+MLP.py - FBCSP + EEGNet + SVM → BCI GDF dataset: loads GDF epochs
Signal processing/FBCSP+EEGNET&SVM.py - Hybrid MLP + XGBoost → BCI GDF dataset: loads GDF epochs
Signal processing/train.py - Game participant → Maze navigation: starts session
Game-Test/bci_maze_game.py - EEG simulation and buffer → Signal processing and prediction: supplies samples
Game-Test/bci_maze_game.py - Signal processing and prediction → Maze navigation: applies predictions
Game-Test/bci_maze_game.py - Maze navigation → Trial metrics: records outcomes
Game-Test/bci_maze_game.py - Trial metrics → Game session logs: writes session logs