1. Clone this project:
   $ git clone https://github.com/diogoshibata/face_recognition_api

2. Go to face_recognition_api folder:
   $ cd face_recognition_api

3. Install TensorFlow and cuDNN:
   $ chmod +x tensorflow-install.sh
   $ ./tensorflow-install.sh
   For cuDNN, look at Step 5 of Dmitriy Kisil's post: https://medium.com/better-programming/install-tensorflow-1-13-on-ubuntu-18-04-with-gpu-support-239b36d29070

4. Install the application:

   • If you have CPU:
     $ chmod +x install-app-cpu.sh
     $ ./install-app-cpu.sh

   • If you have GPU:
     $ chmod +x install-app-gpu.sh
     $ ./install-app-gpu.sh

5. Start the application:
   $ chmod +x start-app.sh
   $ ./start-app.sh

6. Open your browser and access the Console:
   http://<HOST_IP>:5000/ui

It is possible to download or upload all your images to Oracle Cloud Infrastructure Object Storage.

1. Open the deployment.json file and change the following parameter:

   • "OCI_STORAGE_SYNC": true

2. Update the conf/config.prod file with your Oracle Cloud Account information.

3. Generate public and private API keys.

cd  ~/face_recognition_api/conf

openssl genrsa -out ~/face_recognition_api/conf/oci_api_key_private.pem 2048

chmod go-rwx ~/face_recognition_api/conf/oci_api_key_private.pem

openssl rsa -pubout -in ~/face_recognition_api/conf/oci_api_key_private.pem -out ~/face_recognition_api/conf/oci_api_key_public.pem

cat ~/face_recognition_api/conf/oci_api_key_public.pem
-----BEGIN PUBLIC KEY-----
MIIBIjANBgkqhkiG9w0BAQEFAAOCAQ8AMIIBCgKCAQEAo/7hKYCiExQJjo3C4M+V
QVgBowAUYRRRCEIkkfolpYSlHuyfl0Y34OuYlat6mqnZReuSl6MKa2e8PKIHfa4T
0s/01cWvgETYnFty6lskNQO3eGO1KIQ5rZCSuCgcrnWNUzYKQ+0FopSDHj1BPzbv
9Kr13cE9BRU7nBYyG4hGbjUmY9qt6LUKR/MzZZjI5+IAlwpywlYN7X2uwUf30K/c
M/zUpgSZNNiI9lwnJB679pAOHBGdK0JfXEDS6VUAITgzHbNQcFcl6567LCd4U15j
1DjWAEWIObtYvzpX0nqUsn8Is8GoS6eTQBes8Nhf+rUgLSodsP8rGBboWtkDfRBk
9wIDAQAB
-----END PUBLIC KEY-----

4. Copy the Public Key output and paste it under Identity > Users > User Details > API Keys.

Use as your mobile app.

1. /face/add: This operation adds a new human face to the Face Recognition API database. You can upload videos or images that contain only one person.

2. /face/classify: This opertaion to execute a face recognition. You can upload images that contains only one person.

3. /face/train: This operation trains the model to recognize the new faces added by the "/face/add" operation.

4. /face/restart: This operation reloads the Face Recognition API model if you experience any problems.

I use the implementation in this code.

This project is