Facial and Speech Emotion Detection for Stress Monitoring

  • Unique Paper ID: 177475
  • Volume: 11
  • Issue: 12
  • PageNo: 756-759
  • Abstract:
  • Stress is now a common problem in modern life, frequently going unnoticed until it has an impact on one's physical and mental health. This study offers a multimodal stress detection framework that uses facial emotion analysis and speech recognition to more accurately predict stress levels. A refined DistilBERT model is used to analyse speech input into text before classifying emotions and mapping them to appropriate stress levels. In parallel, a Convolutional Neural Network (CNN) trained on the FER dataset is used to identify emotional states and MTCNN is used for face alignment in order to detect facial expressions. In both modalities, the suggested method shows good classification performance and has the potential to be included into real-time stress monitoring applications.

Cite This Article

  • ISSN: 2349-6002
  • Volume: 11
  • Issue: 12
  • PageNo: 756-759

Facial and Speech Emotion Detection for Stress Monitoring

Related Articles