Masked contrastive graph representation learning for age estimation

By Yuntao Shou, Xiangyong Cao, Huan Liu, Deyu Meng. [arXiv link]

This is an official implementation of 'Masked contrastive graph representation learning for age estimation' 🔥. Any problems, please contact shouyuntao@stu.xjtu.edu.cn. Any other interesting papers or codes are welcome. If you find this repository useful to your research or work, it is really appreciated to star this repository ❤️.

🚀 Installation

Pytorch 1.7.0,
timm 0.3.2,
torchprofile 0.0.4,
apex

Training

Morph dataset

python train_MORPH.py --dataset MORPH --lr 0.003 --l2 0.0003 --dropout 0.5 --epochs 100 --w_loss1 1 --w_loss2 1 --w_loss3 1 --margin1 0.8 --margin2 0.2 --NN 4

FGNET dataset

python train_MORPH.py --dataset FGNET --lr 0.0005 --l2 0.00003 --dropout 0.5 --epochs 100 --w_loss1 1 --w_loss2 1 --w_loss3 1 --margin1 0.8 --margin2 0.2 --NN 4

CACD dataset

python train_MORPH.py --dataset CACD --lr 0.001 --l2 0.00003 --dropout 0.5 --epochs 120 --w_loss1 1 --w_loss2 1 --w_loss3 1 --margin1 0.8 --margin2 0.2 --NN 4

If our work is helpful to you, please cite:

@article{SHOU2024110974,
title = {Masked contrastive graph representation learning for age estimation},
journal = {Pattern Recognition},
pages = {110974},
year = {2024},
issn = {0031-3203},
doi = {https://doi.org/10.1016/j.patcog.2024.110974},
url = {https://www.sciencedirect.com/science/article/pii/S0031320324007258},
author = {Yuntao Shou and Xiangyong Cao and Huan Liu and Deyu Meng},
}

Star History