Arcface-pytorch

This is my own implementation for Arcface to be used for deep face recognition, as listed in this paper.\ ArcFace: Additive Angular Margin Loss for Deep Face Recognition

Setting up environment

conda create --name arcface_pytorch --file environment.yml

Setting up training dataset

Download CASIA-WebFace and CACD2000, and extract them inside raw_datasets. The directory should look like this

raw_datasets
│   casia_webface_labels.txt
└───CACD
│   │   14_Aaron_Johnson_0001.jpg
│   │   14_Aaron_Johnson_0002.jpg
│   │   .....
└───CASIA-WebFace
│   └───0000045
│       │   001.jpg
│       │   002.jpg
│       │   ...
│   └───0000099
│       │   001.jpg
│       │   002.jpg
│       │   ...  
│   └───.......

• Make a directory in the root directory called datasets
• Run python pre_processing.py
• Wait until it is done. The result directory dataset should look like this:

  datasets
  │   └───50 Cent
  │       │   001_normal.jpg
  │       │   002_horizontal.jpg
  │       │   ...
  │   └───...

  Training

• Create models directory in project root
• You can change training settings from settings.py
• To start training, run python train.py
• Once training is done, you will see the model in models directory

References

ArcFace: Additive Angular Margin Loss for Deep Face Recognition
A Discriminative Feature Learning Approachfor Deep Face Recognition
Ring loss: Convex Feature Normalization for Face Recognition
InsightFace_Pytorch
focal-loss.pytorch
insightface