Automatic identification of depressive symptoms in college students: an application of deep learning-based CNN (Convolutional Neural Network)
Affiliation:
1. School of Management and Economics , North China University of Water Resources and Electric Power , Zhengzhou , Henan , , China . 2. Mental Health Education Centre, Henan Water Conservancy and Environment Vocational College , Zhengzhou , Henan , , China .
Abstract
Abstract
Facial behavior is the most direct and easily accessible behavioral data. In this paper, based on the face action unit based on FACS, we have conducted quantitative research on the expression behavior pattern of depressed people with digital features through the DAIC-WOZ corpus dataset and E-DAIC dataset and completed the construction of the expression behavior and the application of the automatic identification model of college students’ depressive symptom with the optimization of CNN-LSTM method. For the experimental design and result analysis of the time-frequency ratio of expression behavior and the dynamic rate of change of expression behavior in depressed patients, the digital features are obtained, and the unique expression behavior pattern of depressed patients is argued. The main findings are as follows: Compared to the normal population, depressed patients have special behavioral patterns in emotional feedback and emotional cognition. The characteristics of reduced positive emotional feedback, enhanced negative emotional feedback, easy-to-misjudge neutral stimuli as negative stimuli, and slow changes in expression behavior are mostly indicative of this. By studying the two aspects of the time-frequency ratio of expression behavior and dynamic rate of change of expression behavior, it is found that the CNN-LSTM model obtains 73.21% recognition accuracy and 85.71% recall rate when applied, which is more suitable for depression primary screening scenarios. This paper’s research results offer a methodological basis and technical support for automatically identifying depressive symptoms in college students.
Publisher
Walter de Gruyter GmbH
Reference21 articles.
1. Li, Z., Liu, F., Yang, W., Peng, S., & Zhou, J. (2021). A survey of convolutional neural networks: analysis, applications, and prospects. IEEE transactions on neural networks and learning systems, 33(12), 6999-7019. 2. Gu, J., Wang, Z., Kuen, J., Ma, L., Shahroudy, A., Shuai, B., ... & Chen, T. (2018). Recent advances in convolutional neural networks. Pattern recognition, 77, 354-377. 3. Yamashita, R., Nishio, M., Do, R. K. G., & Togashi, K. (2018). Convolutional neural networks: an overview and application in radiology. Insights into imaging, 9, 611-629. 4. Aghdam, H. H., & Heravi, E. J. (2017). Guide to convolutional neural networks. New York, NY: Springer, 10(978-973), 51. 5. Krichen, M. (2023). Convolutional neural networks: A survey. Computers, 12(8), 151.
|
|