Sign up to save tools and stay up to date with the latest in AI
bg
bg
1

GitHub - nkasmanoff/pi-card: Raspberry Pi Voice Assistant

May 13, 2024 - github.com
The article discusses Pi-card, an AI-powered voice assistant that operates entirely on a Raspberry Pi. It can perform tasks similar to a standard LLM (like ChatGPT) in a conversational setting, and if equipped with a camera, it can take photos, describe what it sees, and answer questions about the image. The system is designed to be completely offline, ensuring user privacy and data protection. It listens for a wake word to start a conversation and can be connected to external APIs or services for enhanced functionality.

To use Pi-card, one needs to download the repository, install the requirements, and follow the setup instructions. The hardware required includes a Raspberry Pi 5 Model B, a USB Microphone, a Speaker, and a Camera. The software setup involves using cpp implementations of audio transcription and vision language models, specifically whipser.cpp and llama.cpp. The hardware setup involves connecting the microphone, speaker, and camera to the Raspberry Pi.

Key takeaways:

  • Pi-card is an AI-powered voice assistant that runs entirely on a Raspberry Pi, capable of conversational tasks and, with a camera, can take and describe photos.
  • The system is designed to be offline, ensuring user privacy and data protection by not sending data to third-party servers.
  • The hardware required for the setup includes a Raspberry Pi 5 Model B, a USB microphone, a speaker, and a camera.
  • The software setup involves using cpp implementations of audio transcription and vision language models, specifically whipser.cpp and llama.cpp.
View Full Article

Comments (0)

Be the first to comment!