“AI Based Facial Emotion Detection”, developed using many machine learning algorithms including convolution neural networks (CNN) for a facial expression recognition task. The goal is to classify each facial image into one of the seven facial emotion categories considered in this study.Trained CNN models with different depth using gray-scale images from the Kaggle website.CNN models are developed in Pytorch and exploited Graphics Processing Unit (GPU) computation in order to expedite the training process. In addition to the networks performing based on raw pixel data,Hybrid feature strategy is employed by which trained a novel CNN model with the combination of raw pixel data and Histogram of Oriented Gradients (HOG) features. To reduce the over fitting of the models,different techniques are utilized including dropout and batch normalization in addition to L2 regularization. Cross validation is applied to determine the optimal hyper-parameters and evaluated the performance of the developed models by looking at their training histories. Visualization of different layers of a network is presented to show what features of a face can be learned by CNN models. Based on the emotion the program recommends the music for the user to up flit the mood.