Facial expression recognition using DenseNet

FER is a field of study that focuses on categorizing human emotions based on their facial expressions. It may be utilized in video games, identifying suspicious people, patient's painful situation at the hospital, online meeting, the E-Iearning system, music player playing songs based on person's mood, driver's tiredness from expressions while driving, robotics, behavioral science, and so on could all benefit from a system that could automatically recognize human emotions from their facial expressions. There are a number of obstacles to overcome while implementing the FER system. The majority of datasets contains photographs of posed humans wearing a specific expression. This is the first hurdle, as real-time applications necessitate a model with no posed or guided emotions. The second issue is that the labels in the datasets are broadly categorized, which implies that in real time, the system may be able to accurately identify some phrases. Here the DenseNet based algorithm may be used to identify real-time facial expressions. Our suggested framework has an AUC of 91.2 percent on the FER-2013 dataset and obtains new state-of-the-art performance.