Identification of Facial Emotions in Hitech Modern Era

At the moment, one of the most significant areas of pattern recognition research is facial expression recognition. This research study proposes a novel method for identifying people's facial expressions based on their moods. Emotions are sudden impulses accompanied by alterations in the facial muscles that result in emotions on the face. The objective of emotion recognition is to identify which of the seven basic emotions-joy, sadness, anger, contempt, surprise, dread, and no-emotion these facial expressions belong to. This topic is difficult due to the complexity of human expressions, which has several applications, including the ability to predict behavior. Deeper architectures have helped researchers understand emotions at the cutting edge of technology. In this research, Convolutional Neural Network, a Deep Learning-based technique, is utilized. Using numerous benchmarked datasets, including FER-2013, CK+ and RAF-DB, the suggested approach's overall efficacy and accuracy are assessed. Experimental results showed that the recommended method obtains an accuracy of 91.12, 92.52%, and 93.83% for Inception-V3, VGG19, and ResNet-50, respectively, on benchmarked datasets as well as our custom-made dataset. ResNet-50 outperforms the competition in real-time learning scenarios with an accuracy of 93.83% for classifying facial emotions.