An eInk picture frame that you can describe the picture you want, and it will display it.
Designed for Inky Impression's 5.7" 7 color eInk display.
It uses Google API speech to text to transform your words into text; it uses OpenAI's Dalle to AI generate the prompt into an iamge.
-
Sign up for Dalle
- Get bearer token, see readme: https://github.com/ezzcodeezzlife/dalle2-in-python
- export OPENAPI_BEARER_TOKEN='sess-d...FdkA'
-
In google cloud console
-
gcloud iam service-accounts create
speech-to-text-quickstart
--project aiframe-364203 -
gcloud projects
add-iam-policy-binding
aiframe-364203 --member
serviceAccount:speech-to-text-quickstart@aiframe-364203.iam.gserviceaccount.com
--role roles/speech.serviceAgent -
gcloud iam service-accounts keys
create speech-to-text-key.json
--iam-account
speech-to-text-quickstart@aiframe-364203.iam.gserviceaccount.com -
export
GOOGLE_APPLICATION_CREDENTIALS=speech-to-text-key.json
-
- run
pipenv install - install pyaudio with:
CFLAGS="-I/opt/homebrew/include -L/opt/homebrew/lib" python3 -m pip install pyaudio- or, maybe:
CFLAGS="-I/opt/homebrew/include -L/opt/homebrew/lib" pipenv install pyaudio
-
Setup ssh:
export RPI_IP={the rpi ip address} -
SSH to RPi
-
curl https://get.pimoroni.com/ink | bash
-
Install some packages:
sudo apt-get install portaudio19-dev(support forpyaudio)
-
Then manually use
pip3to install all the stuff inPipfile(pipenv isnt working on my rpi as of now) -
Deploy (i.e., "copy") software to rpi (from host):
./deploy.sh -
Enable the service on RPi:
sudo cp aiframe.service /etc/systemd/system/sudo systemctl enable aiframe# So it starts on startupsudo systemctl start aiframesudo systemctl status aiframe# Check the statustail -f /var/log/aiframe.log# Tail the logs
-
Add ENV VARS to service:
- See: https://serverfault.com/questions/413397/how-to-set-environment-variable-in-systemd-service
- Run
sudo systemctl edit aiframe - Add something like:
[Service]
Environment="OPENAPI_BEARER_TOKEN=sess-dYH.....hfFdkA"
Environment="GOOGLE_APPLICATION_CREDENTIALS='/home/pi/app/speech-to-text-key.json'"