EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC-ND 4.0

Identification of psychological stress from speech signal using deep learning algorithm

Ankit Kumar, Mohd Akbar Shaun, Brijesh Kumar Chaurasia

e-Prime - Advances in Electrical Engineering, Electronics and Energy · 2024-09 · 9 · 100707

DOI: 10.1016/j.prime.2024.100707 ↗

Kurz gefasst

Fachartikel in der Zeitschrift e-Prime - Advances in Electrical Engineering, Electronics and Energy, 2024-09. Verfasst von Ankit Kumar und 2 weiteren Autor:innen. Die Arbeit wurde bisher 26-mal zitiert (Crossref) und nennt 30 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0). Gefördert von: Kasturba Medical College, Manipal, St. George's, University of London.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Psychological activities have various dimensions in which they correlate with their respective behavior generated by the human body. Understanding the relationship of psychological events with the help of external action units is one of the research subjects to explore various human behavior and their dependencies. Existing work applied various deep learning algorithms to outline the correlation between psychological activities with human emotions. The study of psychological analysis in the medical field is very time-consuming and costly. It requires constant monitoring of the patient for a period of time and various interrogation sessions to finalize the emotional severity of an individual. The challenges in medical expertise over the study of human emotions propels the requirement of computer vision techniques to underscore the recognition of emotions especially in disorder cases. The proposed study explicitly evaluates the recognition of psychological unrest in terms of stress features with the help of speech signals. The study uses a deep learning-based model containing long short-term memory and feed-forward networks to recognize the severity level of psychological illness. The study utilized a standard speech dataset that was collected from various patients against a standard questionnaire format in which a few states of India have been taken to conduct the survey. Patients whose cortisol level goes beyond 10% have been chosen for collecting their speech samples. The responses of each patient have been collected in terms of speech signals to evaluate the correlation between speech and psychological activities. The analysis of the spectrogram of mel-filterbank coefficients of speech signal has been established which further classified into stress and non-stress features. The proposed model achieves 98% of average accuracy for the classification of stress and non-stress features in 150 subjects of the speech dataset. The model is found to be robust for various applications such as preventing suicidal cases, improving decision-making in the diagnosis of depression patients, improves the overall mental healthcare system.

Förderung

  • Kasturba Medical College, Manipal
  • St. George's, University of London

Zitieren

APA

Kumar, A., Shaun, M. A., & Chaurasia, B. K. (2024). Identification of psychological stress from speech signal using deep learning algorithm. e-Prime - Advances in Electrical Engineering, Electronics and Energy, 9, 100707. https://doi.org/10.1016/j.prime.2024.100707

Vancouver

Kumar A, Shaun MA, Chaurasia BK. Identification of psychological stress from speech signal using deep learning algorithm. e-Prime - Advances in Electrical Engineering, Electronics and Energy. 2024;9:100707. doi:10.1016/j.prime.2024.100707

BibTeX
@article{Kumar2024,
  author = {Kumar, Ankit and Shaun, Mohd Akbar and Chaurasia, Brijesh Kumar},
  title = {Identification of psychological stress from speech signal using deep learning algorithm},
  journal = {e-Prime - Advances in Electrical Engineering, Electronics and Energy},
  year = {2024},
  volume = {9},
  pages = {100707},
  doi = {10.1016/j.prime.2024.100707}
}
Kostenlos starten →