Efficient facial emotion recognition model using deep convolutional neural network and modified joint trilateral filter

Facial emotion recognition extracts the human emotions from images and videos. As such, it requires an algorithm to understand model relationships between faces facial expressions recognize emotions. Recently, deep learning models are utilized improve performance of recognition. However, suffer overfitting issue. Moreover, perform poorly for which have poor visibility noise. Therefore, in this paper, efficient learning-based is proposed. Initially, contrast-limited adaptive histogram equalization (CLAHE) applied input images. Thereafter, a modified joint trilateral filter obtained enhanced remove impact impulsive Finally, convolutional neural network designed. Adam optimizer also optimize cost function networks. Experiments conducted by using benchmark dataset competitive models. Comparative analysis demonstrates that proposed performs considerably better compared