TY  - BOOK
AU  - Hannah Yshel A. Dela Cruz, Aaron Joshua M. Lozano, Aubrey Rose B. Marabi, Emmanuel L. Vivero.
AU  - 
ED  - 
ED  - 
ED  - 
ED  - 
SN  - 2
PY  - 4544///446
CY  - 
PB  - 
KW  - 
KW  - 2
KW  - 0
KW  - 6
KW  - 20
N1  - Undergraduate Thesis : { Bachelor of Science in Electronics Engineering) - Pamantasan ng Lungsod ng Maynila, 2024; 5
N2  - ABSTRACT:  Non-verbal behaviors are one of the key factors that determine a person's mental well-being. These behaviors are divided into several characterisitcs, such as body language, gestures, paralinguistics, and facial expressions. Facial expression is one of the most common traits observed in an assessment, as it is responsible for a huge portion of non-verbal communication. This study aims to develop a system that will identify the six basic emotions, namely Joy, Sadness, Anger, Fear, Surprise, Disgust, and Neutral in real-time utilizing a Convolutional Neural Network (CNN) using Rasberry Pi to perform image processing and emotion recognition. The device was built with a module capable of recognizing facial features and landmarks using Python libraries such as PyCharm and OpenCV as part of the image processing. This enabled the device to perform face detection, image preprocessing, and feature extraction to distinguish the emotions shown in the facial expressions of the individual. The emotion recognition software was able to detect seven emotional states in real time: joy, sadness, anger, fear, disgust, surprise, and neutral. The device was also capable of recording the session saved in the Rasberry Pi 4 storage, and generating a report through an emotion log in a TXT format, showing the detected emotions throughout the session, and a radar char in a PNG format, providing the frequency of the emotions detected. The facial emotion detection system that was developed was proven to perform effectively with the help of Keras API. The metrics obtained were an Accuracy of 71.67%. Precision of 77.5%, Recall of 94.2%, and F1-Score of 82.95%. The system was able to detect and predict the correct emotion, especially positive instances
ER  -