Skip to content

Latest commit

 

History

History
27 lines (20 loc) · 876 Bytes

README.md

File metadata and controls

27 lines (20 loc) · 876 Bytes

Frame-Transformer-Network

This repository is the released code and data for our ICMR 2017 paper "Frame-Transformer Emotion Classification Network".

Contact: [email protected]

Requirements

  • Tensorflow
  • Python Numpy

Dataset

Our Emotion6 video dataset and attribution labels for Ekman6 dataset are released here.

Citation

Please cite our paper in your publication if it helps your research:

@inproceedings{gao2017frame,
  title={Frame-Transformer Emotion Classification Network},
  author={Gao, Jiarui and Fu, Yanwei and Jiang, Yu-Gang and Xue, Xiangyang},
  booktitle={Proceedings of the 2017 ACM on International Conference on Multimedia Retrieval},
  pages={78--83},
  year={2017},
  organization={ACM}
}