Sunoyon / face-recognition

Face Recognition REST service using dlib

DLIB Face Recognition Service

A face recognition service using dlib library. This service exposes a REST API which will take two face images and identify whether they are from same person or not.

The service is built over flask-rebar framework and containerization support is provided using dockerfile.

Features:

• Expose dlib face recognition service by REST API
• Expose underline parameters of dlib in the API for both face detections and recognition

Packaging:

The service provides standard production packaging as follows.

• Environment (dev/ prod/ test) configuration management
• Logger configuration
• Error management in HTTP layer
• DTO management
• Containerization with docker
• Unit test support

Pre requisite

python3
build-essential 
cmake

Build (without docker)

$ pip install -r requirements.txt

Run in dev mode (without docker)

$ sh run.dev.sh

Swagger Documentation

http://localhost:5001/v1/apidocs-ui

API Documentation

Run in prod mode (without docker)

$ sh run.prod.sh

Run unit test

$ python -m unittest tests.controllers.recognition
$ python -m unittest tests.services.recognition

Docker

# To build the image and up the container
$ docker-compose up --detach

# To stop the container
$ docker-compose down