Skip to content

Amshra267/Exploratory-Project--Person-Identification

Repository files navigation

Exploratory-Project

Person-Identification

Project proposal

Our Appraoch

We used a pretrained yolo to extract all the possible persons in an input image, then using MTCNN extracted the faces which will pass via a siamese facenet to match the person with the subjects present in gallery database.

To download weights

cd yolo_cnfgs
wget https://pjreddie.com/media/files/yolov3.weights

For requirements

pip install opencv-python numpy
pip install facenet-pytorch

Report

Report in Google docs for gifs

Final-Presentation

Resources