One of the most essential components of an individual’s body is the human face and it acts as the main indicator for the behavioral and the emotional state of the individual face and it's very important for the human to extracting the required input from the human face can be done by using camera directly. The mean of this examination is making facial expression Recognition (FER) conspire by Utilizing the CNN Algorithm and tensor flow to recognize the face by the camera. Facial expression analysis is used in a different way to detect human emotions. There are four types of emotions are recognized: happy, sad, angry, neutral depends on the mood. The playlist itself have the songs in the database, it plays the songs according to the mood detect by the Camera. This research paper is effective because we are using the different algorithm i.e. CNN model which is based on Machine Learning which gives accuracy and reduces the time to recognize the emotions