Face Image Expression Recognition Utilizing Deep Learning Models

In interpersonal relationships, human emotional facial expression recognition (FER) is extremely important. Because people differ significantly in how they exhibit their emotions, automated facial expression identification has always been a difficult issue in practical applications. Designing a highly robust feature that detects the development of FER with a minimal network can traverse the latent emotional features in specific general levels and areas. Recently, many methods for automatically analyzing a person's facial expression have been presented. Still, FER is a difficult task due to the following factors; different postures, facial expressions, illumination conditions, and wearing of scarves/glasses. To address the above-mentioned factors, the Deep Learning (DL) algorithms provide the significant enhancement for FER. In this survey, face image expression recognition utilizing deep learning models is presented. For recognizing the facial expressions using images, the datasets, min-max normalization, feature fusion methods based on DL, and classification based on DL techniques are presented in the taxonomy of this study.