Classification of emotions on images through convolutional neural networks as a method of preventing secondary alexithymia

Abstract

Alexithymia can be defined as the inability to verbalize affective states. One of its main causes lies in the lack of learning of emotions during childhood and can prevail until adulthood. Its identification at an early age can solve problems such as depression and cutting that, in severe cases, can lead to suicide. The present investigation shows the implementation of two convolutional neural networks for the classification of emotions through images.