Automatic Facial Expression Recognition Using Deep Learning

Nonverbal communication is essential in interpersonal interactions, and facial expressions are the primary means. The Facial Expression Recognition system is a technique for assessing a person's emotional state based on their facial expressions. The VGG16 model with Transfer Learning is utilized in the facial expression recognition system that has been suggested. The image processing is the foundation of the proposed system, which also includes acquiring images in real-time. The performance of the suggested technique in terms of recognition will be evaluated depending on whether or not it can recognize a single face or many faces simultaneously. After that, data streaming is carried out, during which the recognized face is shown inside the rectangular boxes using the tensor flow library. After the coordinates of the identified faces have been determined, extracting the necessary characteristics from the original movie is possible. You will need the image scaled down and converted to grayscale to identify the emotion on its face. After the facial expression is identified, the next step is categorizing it. At long last, emotions will be recognized due to the combination of data pertaining to face detection and recognition of facial expressions. The suggested model achieves an accuracy of 88% when applied to the JAFFE database, whereas CK+ achieves 77%.