Facial macro expression analysis

• Main Author: Lim, Jie Xian
• Other Authors: Zheng Jianmin
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2024
• Subjects: Computer and Information Science; Facial macro expression; Deep learning; Ensemble method; Cross platform mobile application
• Online Access: https://hdl.handle.net/10356/175364
• Institution: Nanyang Technological University

Facial Expression Recognition (FER) stands at the forefront of advancing human-computer interaction, enabling machines to interpret human emotions from facial expressions with unprecedented accuracy. This project embarks on the exploration and enhancement of Facial Macro Expression Analysis using cutting-edge machine learning techniques. The core objective is to develop a robust FER system capable of identifying and analyzing a wide range of human emotions with high precision, thereby contributing to the fields of psychology, surveillance, and interactive technology. The methodology employed involves the integration of advanced neural network architectures, including Convolutional Neural Networks (CNNs) to process and analyze facial expressions from static images. This project leverages a comprehensive dataset consisting of diverse facial expressions, annotated with emotion labels, to train and evaluate the FER models. The primary objective is to develop and implement an ensemble method for accurate and reliable macro expression analysis, utilizing server-side processing to handle computationally intensive tasks. This method circumvents the limitations of mobile device processing, leveraging the advantages of cloud computing. The project aims to integrate this advanced analysis into a cross-platform mobile application, thereby democratizing access to sophisticated facial expression analysis for the general public. In summary, the anticipated outcome is a significant enhancement in the accuracy of emotional state detection through facial cues, with broad implications in fields such as education, customer feedback, and mental health. This report details the methodology, challenges, and potential impact of bringing high-level facial expression analysis into everyday mobile use. It lays the groundwork for future innovations in the field, promising a future where machines can understand and respond to human emotions more empathetically and effective.