• Used MTCNN for Face detection
• Extracted the Facial Features using VGGFace(resnet50 model)
• Used CosineDistance to Identify the Facial Features and return the similarity score .

• MTCNN is a state of art model for Face detection and various other stuffs.
• Multi-task Cascaded Convolutional Networks (MTCNN)
• It is known as MultiTasking because it not only detects the face , it returns the BBox Regression as well as Keypoints coordinates in the faces like Left-Eye,Right-eye,Nose,Mouth
• Research Paper : MTCNN Paper
• Github Repo : Github

For Face Detection run :

python face_detection.py

For Face Identification run:

python face_identify.py