Human Facial Emotion Recognition Using Graphical Cascaded Convolutional Neural Network
M. Tamilselvi
Abstract
The study of human emotions through facial expression analysis has garnered a lot of attention lately. In social interactions, facial expressions are essential for conveying one's emotional state. Understanding one's feelings in any circumstance, especially when verbal communication is ineffective, is facilitated by the recognition and analysis of facial expressions. The basic steps in the emotion detection method are covered in this study. It analyses the various approaches to emotion identification in terms of FER methodologies, datasets, emotions, number of emotions, and publication years. It also explores the numerous facial emotion detection strategies, such as machine learning, deep learning, and others, used in literature. One of the most effective, organic, and instantaneous ways for people to communicate their feelings and intentions is through facial expressions. We describe a new approach to fully automatic facial expression identification in this research. Facial landmarks are identified in order to categorise different expressions. A graphical cascaded convolutional neural network (G-CCNN) is suggested for the extraction of features and classification of face expressions. The numerous facial expression databases were used for the tests. The outcome demonstrates that the suggested G-CCNN FER technique can successfully recognise objects up to 93.21%, 95.69%, and 98.7 for FER2013, RAF-DB, and CK+, respectively.