Tools
Tools
Ponugupati, Siva Naga Malleswara Rao (2024) A Multimodal system for detecting life threatening emotions in social media using AI. Masters thesis, Dublin, National College of Ireland.
Preview |
PDF (Master of Science)
Download (1MB) | Preview |
Preview |
PDF (Configuration Manual)
Download (544kB) | Preview |
Abstract
Catastrophic emotions like anxiety, fear, anger, and suicidal phrases are more shortly experienced as a result of social and psychological issues occurring after COVID-19. With the tremendous amount of content in various formats on social media, we present a multimodal AI system that detects indicators of early stages of the emotional distress based on speech, vision and text analysis. It uses the latest deep learning models, spectrogram-based LSTMs for speech, facial recognition for vision and sentiment analysis for text to improve detection and timely report generation. As a result of filling the gaps left by other single-modal approaches, our framework can provide timely interventions for these patients, help mental health professionals, and raise the quality of patients’ treatment. Its scale and its ethical structure provides a strong framework for delivering intelligent analytics into healthcare, thereby gaining lives of those in danger and allowing for mental health crisis intervention. The findings of this study also show how AI can be used to revolutionize other mental health issues across the world.
Actions (login required)
 |
View Item |