Real-time face recognition in unconstrained environments, based on InsightFace.
install requirements
pip install -r requirements.txt
For face detection, run the following command:
python inference_detection.py --input input/sajjad.jpg
For face identification, feature extraction and compare, run the following command:
python inference_identification.py --input1 input/sajjad0.jpg --input2 input/sajjad1.jpg
For face recognition, run the following command:
python inference_recognition.py --input ./io/input/sajjad2.jpg --update
The output will be saved in ./output
by default.
If you want recognize your own faces, first you should put your images in ./face_bank
directory with your names as directories name in this structure:
face_bank
│
└───Ali
│ │ ali_image_1.jpg
│ │ ali_image_2.jpg
│ │ ...
│
└───Sara
│ │ sara_image_1.jpg
│ │ sara_image_2.jpg
│ │ ...
│
│
...
Then run same command as above with --update
argument. Note that After each change in ./face_bank
directory, you should use --update
again.