search

Cluoyao / Siamese-network-based-on-face-recognition

Siamese network application (by Resnet-50)
14 stars 1 forks source link

readme

Siamese-network-based-on-face-recognition

Siamese network application (by Resnet-50)

Requirements

.Python 3.7
.Pytorch 1.0

Training Dataset

BaiduYun: https://pan.baidu.com/s/19QNGst2E1pRKL7Dtx_L1MA
Google Drive: https://drive.google.com/file/d/17LfvBZFAeXt0ACPnVckfdrLTMHUpIQqE/view
@misc{3ddfa_cleardusk,
author = {Jianzhu Guo, Xiangyu Zhu and Zhen Lei},
title = {3DDFA},
howpublished = {\url{https://github.com/cleardusk/3DDFA}},
year = {2018}
}

Test Dataset

BaiduYun: https://pan.baidu.com/s/1DTVGCG5k0jjjhOc8GcSLOw
Google Drive: https://drive.google.com/file/d/1r_ciJ1M0BSRTwndIBt42GlPFRv6CvvEP/view
@misc{3ddfa_cleardusk,
author = {Jianzhu Guo, Xiangyu Zhu and Zhen Lei},
title = {3DDFA},
howpublished = {\url{https://github.com/cleardusk/3DDFA}},
year = {2018}
}

Experiments

Trained a model with default configuration (Restnet-50 for 50 epochs). The model can be downloaded from BaiduYun keywords: atmq
Model accuracy is: 0.9730000000000001
Results shown as follow (ROC Curve):
image

TODO: Train the Siamese-network with CASIA datasets.
TODO: Add usages