Scholarlink

A rapid hike in interest has been found in the realm of facial exprerssion analysis for finding the human emotions and anticipating mental health. In order to communicate one’s emotions to others, these facial expressions playa pivotal role and can contribute in providing valuable insights and perspectives regarding individual’s mental health. By making advantage of deep learning techniques, ESCNN (Enhanced Stress Convolutional Nueral Network) proves to be a vigorous and efficient approach for mental health prediction. This ESCNN precissely findout a range of facial expressions, dividing them into different emotional states like anger, disgust, neutrality, surpise, happy, sadness and fear. The methodology includes transfer learning with MobileNet V and TensorFlow, making use of pre-trained data from the Cohn-Kanade+ dataset for Facial Expression Recognition (FER). More significantly as evidenced by the comprehensive experimental analysis in ESCNN demonstrating higher grade performance in tasks which involve stress recognition and mental heakth prediction through the integration of transfer of learning with Haar Cascade face detection, thereby remarkable prevailing methodologies.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Facial Expression Recognition System Using Enhanced Stress Convolutional Neural Network for Mental Health Prediction

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Haitham Alhussain,
  • Anwar Ahmed Alabdulathem,
  • Vemparala Priyatha,
  • Mohammed Bashit,
  • Deepak Hajoary,
  • V. Parimyndhan

摘要

A rapid hike in interest has been found in the realm of facial exprerssion analysis for finding the human emotions and anticipating mental health. In order to communicate one’s emotions to others, these facial expressions playa pivotal role and can contribute in providing valuable insights and perspectives regarding individual’s mental health. By making advantage of deep learning techniques, ESCNN (Enhanced Stress Convolutional Nueral Network) proves to be a vigorous and efficient approach for mental health prediction. This ESCNN precissely findout a range of facial expressions, dividing them into different emotional states like anger, disgust, neutrality, surpise, happy, sadness and fear. The methodology includes transfer learning with MobileNet V and TensorFlow, making use of pre-trained data from the Cohn-Kanade+ dataset for Facial Expression Recognition (FER). More significantly as evidenced by the comprehensive experimental analysis in ESCNN demonstrating higher grade performance in tasks which involve stress recognition and mental heakth prediction through the integration of transfer of learning with Haar Cascade face detection, thereby remarkable prevailing methodologies.