MINDGAN BCI Suite

A desktop application that ties the whole motor-imagery BCI workflow together — real-time EEG inference, training, and drone control.

MINDGAN BCI Suite is a desktop application I built to tie my motor-imagery BCI research together — one program for the whole loop instead of a folder of loose scripts. It is built around the Emotiv EPOC X headset.

  • Real-time inference engine that reads EEG from LSL streams or the Emotiv Cortex API
  • The MINDGAN training pipeline (cDCGAN curriculum) with automatic figure export
  • DJI Tello drone control driven by motor-imagery classification
  • Six modules covering recording, preprocessing, training, and online testing

The models inside are the same ones from my MINDGAN paper.

Stack: Python, PyTorch, Tkinter, LSL, Emotiv Cortex API

Code: github.com/Meshkat22/MINDGAN_BCI_SUITE