This is a model repository for the following article:
Grm, K., Štruc, V., Artiges, A., Caron, M., Ekenel, H. K. Strengths and Weaknesses of Deep Learning Models for Face Recognition Against Image Degradations. Published in: IET Biometrics, 2017.
The article has been published at the IET digital library. An arXiv copy is also available.
We provide the pre-trained face recognition models for four common CNN architectures, namely,
- AlexNet
- GoogLeNet (InceptionV3)
- VGG-Face
- SqueezeNet
The models were trained on a subset of the VGG face dataset that contains approximately 1.8 million face images across 2622 identities.
Model definitions and weights are available here.
The following software versions were used and are recommended:
- Python 2.7
- Numpy 1.13
- h5py 2.7
- Theano 0.8
- Keras 1.2
- OpenCV 2.4
If you use our trained models, please cite the above paper as follows:
@article{grm2017strengths,
title={Strengths and weaknesses of deep learning models for face recognition against image degradations},
author={Grm, Klemen and {\v{S}}truc, Vitomir and Artiges, Anais and Caron, Matthieu and Ekenel, Haz{\i}m K},
journal={IET Biometrics},
volume={7},
number={1},
pages={81--89},
year={2017},
publisher={IET}
}