Inspired by the project detailed in this DFRobot community post, the Unihiker AI Assistant aims to enhance the functionalities provided by the original implementation. You can check out the original repository here.
The primary goal of this project is to enable local speech-to-text and text-to-speech functionalities. Initially, we explored using sherpa-onnx, which appeared promising. However, we discovered another intriguing library that better suits our needs:
A fast, local neural text-to-speech system that sounds great and is optimized for the Raspberry Pi 4.
We have integrated some of Piper's capabilities into this project to enhance its performance and usability.
- Local Speech-to-Text: Efficiently converts spoken language into text without relying on external servers.
- Local Text-to-Speech: Converts text back into natural-sounding speech using the Piper library.
- Optimized for Raspberry Pi 4: Ensures smooth performance on low-resource hardware.
To get started with the Unihiker AI Assistant, follow these steps:
-
Clone this repository:
git clone https://github.com/sofianhw/ai-unihiker.git
-
Navigate to the project directory:
cd ai-unihiker -
Install the required dependencies:
pip install -r requirements.txt
After installation, you can start the AI assistant by running:
python main.py Python3.9
This project is licensed under the MIT License.
Feel free to reach out with any questions or feedback. Happy coding!