A basic Flask based face recognition API using face_recognition and PostgreSQL

Based on the following repo: https://github.com/vearutop/face-postgre

The API exposes two POST endpoints:

• faces which adds a face to the database and the remote URL of the image file
• faces/searches which finds a face in the database

• Follow the instructions for face_recognition installation so you can get dlib installed
• pip3 -r requirements.txt
• You'll need docker and docker compose on your machine to run the PostgreSQL database (or just swap for a local instance)

mkdir uploads # needed to store the files temporarily
docker-compose up
python3 setup-db.py # needed to setup PostgreSQL properly
DATABASE_STRING="pq://user:pass@localhost:5434/db" python3 app.py

The docker image takes some time as dlib needs to be built manually and then run the docker compose set up by removing the commented lines

In the docker compose are 2 services:

• db which is the PostgreSQL database
• api which is the API server itself

docker build --tag face_rec_api .
docker-compose up
python3 setup-db.py # needed to setup PostgreSQL properly

issue currently with Flask saying that POST methods aren't allowed