Deep emotion recognition has a wide range of applications, including human-robot communication, business, movies, services hotels, and even politics. Despite the use of various supervised and unsupervised methods in many different fields, there is still a lack of accurate analysis. Therefore, we have taken on this challenge as our research problem. We have proposed a mechanism for efficient and fine-grained classification of human deep emotions that can be applied to many other problems in daily life. This study aims to explore the best-suited algorithm along with optimal parameters to provide a solution for an efficient emotion detection machine learning system. In this study, we aimed to recognize emotions from facial expressions using deep learning techniques and the JAFFE dataset. The performance of three different models, a CNN (Convolutional Neural Network), an ANN (Artificial Neural Network), and an SVM (Support Vector Machine) were evaluated using precision, recall, F1-score, and accuracy as the evaluation metrics. The results of the experiments show that all three models performed well in recognizing emotions from facial expressions. The CNN model achieved a precision of 0.653, recall of 0.561, F1-score of 0.567, and accuracy of 0.62. The ANN model achieved a precision of 0.623, recall of 0.542, F1-score of 0.542, and accuracy of 0.59. The SVM model achieved a precision of 0.643, recall of 0.559, F1-score of 0.545, and accuracy of 0.6. Overall, the results of the study indicate that deep learning techniques can be effectively used for recognizing emotions from facial expressions using the JAFFE dataset.