Combining Sequential Geometry and Texture Features for Distinguishing Genuine and Deceptive Emotions
Abstract
In this paper, we explore a new type of automatic emotion recognition task - distinguishing genuine and deceptive emotions from video clips. For this task, it is not enough only using static images clipped from the video data, as there's only subtle differences between two types of emotions, which makes it even harder for automatic analysis. To utilize the temporal information, we introduce temporal attention gated model for this emotion recognition task. Compared to texture features which describe the whole face area, the facial landmark sequences may also indicate the temporal changes of the face, thus we utilize them by encoding feature sequence unsupervisedly.
Cite
Text
Li et al. "Combining Sequential Geometry and Texture Features for Distinguishing Genuine and Deceptive Emotions." IEEE/CVF International Conference on Computer Vision Workshops, 2017. doi:10.1109/ICCVW.2017.372Markdown
[Li et al. "Combining Sequential Geometry and Texture Features for Distinguishing Genuine and Deceptive Emotions." IEEE/CVF International Conference on Computer Vision Workshops, 2017.](https://mlanthology.org/iccvw/2017/li2017iccvw-combining/) doi:10.1109/ICCVW.2017.372BibTeX
@inproceedings{li2017iccvw-combining,
title = {{Combining Sequential Geometry and Texture Features for Distinguishing Genuine and Deceptive Emotions}},
author = {Li, Liandong and Baltrusaitis, Tadas and Sun, Bo and Morency, Louis-Philippe},
booktitle = {IEEE/CVF International Conference on Computer Vision Workshops},
year = {2017},
pages = {3147-3153},
doi = {10.1109/ICCVW.2017.372},
url = {https://mlanthology.org/iccvw/2017/li2017iccvw-combining/}
}