https://github.com/jojasadventure/hey-aria

Implementation of openwakeword-simplified with a custom wake word
https://github.com/jojasadventure/hey-aria

Last synced: about 1 month ago
JSON representation

Implementation of openwakeword-simplified with a custom wake word

• Host: GitHub
• URL: https://github.com/jojasadventure/hey-aria
• Owner: jojasadventure
• License: other
• Created: 2024-08-29T00:01:05.000Z (5 months ago)
• Default Branch: main
• Last Pushed: 2024-08-29T00:12:38.000Z (5 months ago)
• Last Synced: 2024-11-06T12:09:42.286Z (3 months ago)
• Language: Python
• Homepage:
• Size: 2.83 MB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# Wake Word Detection App

A Python application for real-time wake word detection using customizable ONNX models.

Made possible by the library for https://github.com/sujitvasanth/openwakeword-simplified

Note: Experimental, just for myself, no requirements.txt file at this time, only ran on intel mac from 2015 ( - ; 

## Features

- Real-time audio processing

- Customizable wake word models (ONNX format)

- User-friendly GUI built with customtkinter

- Support for multiple audio devices

- Visual feedback with RMS meter and wake word indicator

## Requirements

- Python 3.7+

- PyAudio

- NumPy

- ONNX Runtime

- customtkinter

## Installation

1. Clone the repository:

   ```

   git clone https://github.com/yourusername/wake-word-detection-app.git

   ```

2. Install the required packages:

   ```

   pip install -r requirements.txt

   ```

## Usage

Run the application:

```

python main.py

```

Select an audio device and model from the dropdowns, then click "Start Listening" to begin wake word detection.

## Adding Custom Models

Place your custom ONNX models in the `models/` directory. They will automatically appear in the model selection dropdown.

## License

[MIT License](LICENSE)