Wavelet 🌊🧠

Wavelet is a transformer based custom model that enables real-time EEG to AI-Derived MEG brain waves enhancing.

Features

• Real-time Signal Processing: Visualize EEG signals with milliseconds latency
• 3D Brain Visualization: Interactive 3D model with real-time sensor activity mapping
• AI Signal Enhancement: Transform EEG signals into MEG-like data for deeper insights
• Frequency Band Analysis: Live monitoring of Delta, Theta, Alpha, Beta, and Gamma bands
• Audio Synthesis: Convert brain signals into real-time auditory feedback (just a cool idea, no use, must try!)
• Data Export: Download raw EEG or enhanced MEG data in CSV/JSON formats

Technology Stack

• Frontend: Next.js 14, TypeScript, Tailwind CSS
• 3D Rendering: Three.js with custom WebGL shaders
• Data Visualization: Recharts with custom animations
• Signal Processing: Custom FFT implementation with real-time filtering
• Audio: Web Audio API with dynamic synthesis

Getting Started

1. Clone the repository:

npm run dev
# or
yarn dev
# or
pnpm dev
# or
bun dev

Open http://localhost:3000 with your browser to see the result.

Some Credits and mentions

• This data was obtained from the OpenfMRI database. Its accession number is ds000117.
• Perri Karyal (her work on EEG gaming is incredible,check it out) and Charles Cumpsty for lending us the EEG equipment!
• Tim and Gleb for letting us explore more their initial idea: BrainTransform ❤️
• Nebius for offering the Nvidia H100s used!