Smart-ESP System for Emotion Strength Prediction for Static Facial Images

The detection of a person present emotional state using facial expressions is a useful study subject in facial emotion identification. The difficult problem of automatically recognizing facial emotions is currently a topic of intense research for scholars. One of the main challenges for the face emotion detection system is predicting the intensity of the emotion, and deep learning requires large amounts of data for prediction. The transfer learning approach is mostly employed in prediction systems when there is insufficient data available for training. Our suggestion was to employ this method for determining the intensity of facial emotions.The shared parameters and priors are learned in this study using the parameter transfer learning technique. ResNet-50 The foundational design is built on Convolutional Neural Networks.