Mental Health Assessment: Analyzing Body Language Patterns and Emotional Expression

Abstract

This research focuses on enhancing the understanding of human emotions, particularly for individuals with mental health challenges, by analyzing body language and facial expressions. We aim to develop a real-time, multi-modal emotion recognition system utilizing computer vision and deep learning techniques. Despite recent advancements, existing solutions often lack accuracy in emotion recognition. Our approach integrates various data types and employs deep learning architectures for more precise assessments. We implemented two deep learning models: a ResNet-50 architecture achieved an accuracy of 94.5%, and a custom convolutional neural network reached an accuracy of 94.33%. This framework aims to improve the recognition of body language patterns, facilitating personalized care in mental health monitoring.