We present our preliminary work to determine if patient's vocal acoustic, linguistic, and facial patterns could predict clinical ratings of depression severity, namely Patient Health Questionnaire depression scale (PHQ-8). We proposed a multi modal fusion model that combines three different modalities: audio, video , and text features. By training over AVEC 2017 data set, our proposed model outperforms each single modality prediction model, and surpasses the data set baseline with ice margin.
@article{arxiv.1711.11155,
title = {Predicting Depression Severity by Multi-Modal Feature Engineering and Fusion},
author = {Aven Samareh and Yan Jin and Zhangyang Wang and Xiangyu Chang and Shuai Huang},
journal= {arXiv preprint arXiv:1711.11155},
year = {2017}
}
Comments
Thirty-Second AAAI Conference on Artificial Intelligence (AAAI-18)