Face recognition and identification of specific object is one of the key research areas for assorted domains including forensic applications whereby the suspicious persons or objects can be identified using their live features, behavior and traits. There are many segments in the human face which can be trained and further analyzed for the recognition in forensic applications. These objects are lips, forehead, cheeks, chin and many others which overall make the human smile and moves ahead to the face smile detection. In other areas of research, the work on hairstyle can be done but these can be manipulated, therefore the work on face smile detection is very prominent. In this research work, the deep learning based approach of Convolutional Neural Network (CNN) with the fuzzy logic is presented so that the higher degree of accuracy in the face smile can be done.