cd .config/autostart touch kiezbot.desktop nano kiezbot.desktop
[Desktop Entry] Type=Application Name=kiezbot Exec=cool-retro-term --fullscreen -e /home/pi/Code/kiezbot/start.sh
python3 -m pip install -r requirements.txt
cd .local/share/applications touch kiezbot.desktop nano kiezbot.desktop
[Desktop Entry] Type=Application Name=kiezbot Terminal=false Type=Application Categories=Application Exec=cool-retro-term --fullscreen -e /home/pi/Code/kiezbot/start.sh
sudo apt install mpg123
(mp3 player for Linux)
sudo apt-get install python-rpi.gpio python3-rpi.gpio
gpio lib
Our little Kiezbot Berlin is a conversational AI robot simulating three different characters. It's made to interact with the citizens of Berlin to explain AI in a playfull way.
We use a Python script to send a voice recording of user to multiple Artificial Intelligence (AI) APIs:
-
Whisper AI for speech-to-text
-
ChatGPT 3.5 as LLM that generates an answer
-
Google Text-To-Speech to let the answer sound like a real character
To install the requirements (globally) run:
pip install -r requirements.txt
to install missing requirements
Create a .env
file and add your API Keys like this:
OPENAI_API_KEY="your-key"
Run: python main.py
That's it!
Made by
|
A project by
|
Supported by
|