This model, introduced in a paper published in Mobile Networks and Applications, was trained to recognize emotions in human speech by analyzing different relevant features.