Sharma, Sugandha (2023) Facial Emotion and Behavioural Recognition from Video using Deep Learning Approach. Masters thesis, Dublin, National College of Ireland.
Preview |
PDF (Master of Science)
Download (3MB) | Preview |
Preview |
PDF (Configuration Manual)
Download (1MB) | Preview |
Abstract
Security cameras have become essential now a days in public areas or high-risk areas for the security concerns. To ensure this it is important to capture the emotions and actions of the individuals to take an appropriate action in case of any unexpected events. This analysis will help to understand the behavioural characteristics of an individual to prevent crimes. In addition, face detection is implemented in case of tracking purpose. The primary objective of this research is to analyse emotions, actions, and face on video dataset and in real time scenario to understand the behavioural characteristics of individual. Several models like 3D CNN model, RNN, pretrained model like Deep Face, Haar Cascade ClassiTier, MTCNN to predict the emotions and actions by individuals. The study emphasizes on importance of these models in real time scenarios which is a contribution in security sector. It highlights the technical aspects and the role of technology in enhancing the security systems. The performance of classiTier has improved with the large dataset. When models are applied to real-time settings as compared to pre-trained datasets it tens to yield better results.
Actions (login required)
![]() |
View Item |