Facial Expression Recognition using Convolutional Neural Network
Abstract- Enhancing modern-day machines or computer systems for recognizing different facial expressions and also to know about the emotions of humans from them in real-world scenarios is a demanding and challenging area of research. Facial expression recognition (FER) explains systems how recognizing emotions is considered one of the major applications in the field of analyzing patterns and artificial intelligence. The expressions of faces which is happy, sad, angry, surprise, fear, and disgusted are identical across culture. The main goal of this project work is to build a Facial Expression Recognition System to recognize various human emotions. The system is based on the concept of a very well-known Convolution Neural Network (CNN). The live video stream are recorded using the webcam present in the computer systems or which is given as an input for facial feature extractions and is then provided to the network for classification into the seven basic human emotions which finally is the main goal of this research work. The model is made up of various activation layers and each of them is created to undergo various training techniques. The paper also discusses the results of this research work stating the techniques to improve the performance of the model.
Keywords - Convolutional Neural Network (CNN), Emotion Detection, Validation Accuracy, Training methods, Human Facial Expressions.