About

This is a small example for using the face_recognition libary to identify faces in a picutre. For aboiding to intall a complete environment I creating a docker container and use it as a environment.

Build the container

First Generate Container

docker build -t train_image

Run the Image

docker run -v ${PWD}/images:/app/input -v ${PWD}/test:/app/test -v ${PWD}/result:/app/output train_image  

Folderstructure

input

Here are the images that will be idenfitied. You can cluster it into separate subfolders. The filenames will be used as label that are identified.

test

This folder will contains all images that must run throught the identify process. So this will be used to identify the faces

output

When the engine identify a face, it will result the modified image (marked with a bounding box) into this folder. It will use the same filename like the input filename