Real-Time Convolution Neural Network for Emotion Classification

Abstract

Humans communicate mostly through their emotions. Today, as online interviews and classes become more common, it is critical that any channels of communication do not become obstructive. Body language and nonverbal communication play a big role in determining how to interpret a situation and, hence, how to behave. We’ve developed over time, relying significantly on these nonverbal pieces of data to socialize.Real-time emotion classification using a deep neural network is proposed in this research. The framework for developing Convolutional Neural Networks (CNN) is employed. To reduce processing costs, the collected dataset is translated to a suitable format and pixel values are normalized. The success of emotion detection is determined by a high-quality dataset, pre-processing processes, and contemporary CNN architectures that close the gaps between desired and tested accuracies.The proposed model is proven by constructing a video conferencing system that uses CNN architecture to fulfil the tasks of face identification and emotion categorization in tandem. On the JAFEE, FER-2013, and own datasets, accuracy of up to 94 percent is attained.