This project is for classifying the biometric feature in a group. It can control the permission of entrance, which is more secure than the key. For instance, recognize family member to entry home or employees in a company, etc. It will process in this procedure.

1. collect the person’s facial pictures and sort it in specific file structure.
2. detecte the boxes of face
3. dlib shape_predictor_68_face_landmarks will get the facial feature
4. compute features in training data to an 128D vector
5. output to feature fusion

use deep-speaker cnn model to get the feature of voice pass

use face_descript_compute.py to compute facial description

flowchart LR
A(collect image) -->|cnn_face_detection| B(128D vector)
B --> C(feature fusion)

project
│
└───data
│   │
│   └───person_1
│       │   person_1_face-1.jpg
│       │   person_1_face-2.jpg
│       │   ...
│       │   person_1_face-n.jpg
│   │  
│   └───person_2
│       │   person_2_face-1.jpg
│       │   person_2_face-2.jpg
│       │   ...
│       │   person_2_face-n.jpg
│   ...
│   ...
│   │
│   └───person_n
│       │   person_n_face-1.jpg
│       │   person_n_face-2.jpg
│       │   ...
│       │   person_n_face-n.jpg
│   
└───test
    │   file1.jpg
    │   file2.jpg