Facial Emotion Recognition With Sentiment Analysis From Real-Time Image: A Deep Learning Approach

Abstract

Deep learning and machine learning have become more important in current technologies. AI-based computer technology is used for facial detection, identification, and the detection of facial emotions. "Deep Learning" is a subfield in machine learning (DL). It takes a variety of techniques to track and categorize the human face. In computer vision tests, deep learning algorithms have been beating conventional approaches. In many different sectors, applications for facial recognition and detection are utilized to provide identity, security, and verification. Facial emotion recognition makes it easier to determine a person's feelings based on their facial expressions. I focused on successfully identifying faces, detecting human faces, and categorizing facial emotions using sentiment analysis in this paper. To properly recognize and categorize facial expressions using deep learning techniques. I also divided our work into three parts: first, we will accurately detect faces in real-time by using the Haar-Cascade detection algorithm in the first segment; second, the input from the first part will be processed based on the features of the faces as well as the dataset we used in the first segment (CNN, KNN, and RF). In the last section, the face is verified so that the various facial expressions—such as happy, sad, disgusted, surprised, angry, and neutral—can be classified. The goal of the proposed study is to make face detection, recognition, and emotion recognition simpler. I wanted to make sure that the automatic face expression detection and identification algorithms functioned with dignity and accuracy