mobile-face

try on face recognition with mobile deep learning architecture

Requirements

1. pytorch 0.3.0
2. opencv3-python
3. dlib 0.19.0
4. numpy
5. pandas

Current architecture

1. sphereface
2. resnet
3. mobilenetv2
4. angle loss
5. AM loss (implmeanted on Pytorch)

Reference

1. code

• sphereface
• AM softmax
• AM softmax tensorflow
• MTCNN

2. paper

• sphereface
• AM softmax
• resnet
• mobilenetv2
• MTCNN

Thanks to the carwin, this code is heavily based on sphereface_pytorch