Attribute-Aware Convolutional Neural Networks for Facial Beauty Prediction

Attribute-Aware Convolutional Neural Network (AaConv-Net) is a dynamic CNN architecture to integrate the aesthetic-related attribute information that promotes the performance of facial beauty prediction finally.

Versions

There are two options for implementing AaConv-Net, including Caffe and Pytorch versions. BTW, Pytorch version is coming soon...

Citation

Please cite our paper:

@article{lin2019regression,
  title={Attribute-Aware Convolutional Neural Networks for Facial Beauty Prediction},
  author={Lin, Luojun and Liang, Lingyu and Jin, Lianwen and Chen, Weijie},
  journal={International Joint Conference on Artificial Intelligence},
  year={2019}
}

Contact Us

For any questions, please feel free to contact Dr. Lin ([email protected]).

Copyright

This code is free to the academic community for research purpose only. For commercial purpose usage, please contact Prof. Lianwen Jin ([email protected]).