Real Time Emotion Analysis (RTEA)

Recognizing human emotion in real-time is one of the most challenging and powerful tasks in telepsychology. Neural network-based emotion recognition gives a better performance than simple image processing. This project presents the design of a deep learning system that is capable of detecting human emotion through facial and speech emotion recognition. This paper proposes a CNN or convolution neural network-based deep learning. It also discusses the application of human emotion recognition for the purpose of telepsychology. Mental health professionals are provided with real-time emotional data of their patients for better treatment. For the purpose of the project, two datasets are used. One is for facial emotion recognition called AffectNet Database and the other is The Ryerson Audio-Visual Database of Emotional Speech and Song (RAVDESS) for speech emotion recognition. The accuracies achieved with the proposed model are 63 and 77 percent, respectively.