/PersEmoN

PersEmoN: A Deep Network for Joint Analysis of Apparent Personality, Emotion and Their Relationship

Primary LanguageC++

PersEmoN: A Deep Network for Joint Analysis of Apparent Personality, Emotion and Their Relationship

This repository contains the training prototxt for our papers:

PersEmoN: A Deep Network for Joint Analysis of Apparent Personality, Emotion and Their Relationship.
Le Zhang, Songyou Peng, Stefan Winkler
IEEE Transactions on Affective Computing, 2019.

and

Give Me One Portrait Image, I Will Tell You Your Emotion and Personality.
Songyou Peng, Le Zhang, Stefan Winkler, Marianne Winslett
ACM Multimedia, 2018, Demo.

image

Requirements

  • MTCNN
    We use the MTCNN to first detect and align the faces. We used two customized layers which may not be included in the official caffe.
  • TSN
    Used in Videodata layer.

Also, For the "DomainConfusionInnerProduct" layer, we get the code from the following paper: "Simultaneous Deep Transfer Across Domains and Tasks.", ICCV, 2015.

As the original code for the above paper is not well-maintained, we provide the source code of "DomainConfusionInnerProduct" layer in this repository.

Citations

Please cite the following papers if you use this repository in your research work:

@article{zhang2019persemon,
  title={PersEmoN: A Deep Network for Joint Analysis of Apparent Personality, Emotion and Their Relationship},
  author={Zhang, Le and Peng, Songyou and Winkler, Stefan},
  journal={IEEE Transactions on Affective Computing},
  year={2019},
  publisher={IEEE}
}

and

@inproceedings{peng2018mm,
 title = {Give Me One Portrait Image, I Will Tell You Your Emotion and Personality},
 author =  {Peng, Songyou and Zhang, Le and Winkler, Stefan},
 booktitle = {ACM International Conference on Multimedia (ACM MM)},
 year = {2018},
}

Contact Le Zhang ✉️ for questions, comments and reporting bugs.