Skip to content

SanMog/NeuroBridge

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

9 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation


🧠 NeuroBridge v1.0 — Cognitive I/O Platform

NeuroBridge — is a minimalist, high-performance interface designed for systems architects and power users to bridge the gap between human thought and digital execution. It expands the bandwidth of human-computer interaction by replacing manual typing with low-latency neural transcription.

AI-Powered Python 3.12 NVIDIA CUDA License MIT

"The fingers are the bottleneck of the mind. NeuroBridge expands that bandwidth."


✨ Key Features

  • 🎙️ Neural Voice Input: Instant voice-to-cursor transcription using Faster-Whisper (large-v3-turbo).
  • 🔊 High-Speed Auditory Bridge: Rapid content ingestion via Microsoft Edge Neural TTS at 2x+ speed.
  • 🛡️ Privacy First: 100% local voice processing on your GPU. Your technical thoughts never leave your machine.
  • ⚙️ Zero-UI Interaction: Operates as an invisible layer of the OS via global hotkeys. No windows, no clutter.

🏛 Philosophy: Bandwidth Expansion

NeuroBridge is built on the premise that we can think and read much faster than we can type.

  • Fast Data Dump (Input): Offload complex thoughts directly into your IDE/Terminal without breaking the state of "Flow".
  • Rapid Ingestion (Output): Consume documentation through high-fidelity neural voices while maintaining visual focus on the code.

🛡 Engineering & Privacy

NeuroBridge employs a hybrid architecture to balance absolute privacy with high-quality synthesis:

  1. Sovereign Local Input (100% Secure): Speech-to-Text (STT) is handled entirely on your local GPU. Zero-telemetry.
  2. Edge-Neural Output: Text-to-Speech (TTS) utilizes Microsoft’s advanced neural engines via the Edge API for superior phonetic clarity.
  3. Zero-UI: The system lives in your system tray and is controlled via global hotkeys.

⚠️ Known Limitations (v1.0)

  • Language Switching: Language detection is performed on the entire text block. Seamless on-the-fly switching is a roadmap feature for v2.0.
  • Phonetic Accuracy: In mixed-language blocks, the primary language's voice is used, which may result in accented pronunciation of foreign technical terms.

🛠 Technical Stack

  • Runtime: Python 3.12 (Optimized for Windows 10/11)
  • ML Core: CTranslate2 / Faster-Whisper (GPU-bound)
  • Acceleration: CUDA 12.x / cuDNN (Requires NVIDIA RTX 30/40 series)
  • I/O Logic: Win32 API, pystray for state management, edge-tts synchronization.

🚀 Installation

  1. Clone the repository:

    git clone https://github.com/SanMog/NeuroBridge.git
    cd NeuroBridge
  2. Initialize environment:

    python -m venv .venv
    .venv\Scripts\activate
    pip install -r requirements.txt
  3. Execution: Run NeuroLauncher.vbs to initialize the bridge as a background service.


🕹 Usage Protocol

Command Action System Response
Record Hold F4 Speech is typed instantly at cursor position.
Listen Ctrl + Q Clipboard is read aloud via Edge Neural HUD.
Manage Right-Click Tray Restart engines or exit system safely.

Architect: SanMog
Status: 🟢 Operational | v1.1
License: MIT

"The system lives where your focus is."


About

A minimalist, local-first voice interface for high-bandwidth engineering. Powered by Faster-Whisper and Edge Neural Engine.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors