Master's Degree in Telecommunication Engineering.

University of Seville.

Automatic detection of emotions through Electrocardiogram analysis.

Emotions play a big role in human life and constitue an an extremely important aspect in interpersonal communication and interaction, and nowadays, in a hyper-connected world in which society is increasingly dependent on technology, most of human-computer interaction systems are still deficient when it comes to identifying the emotional states of people.

Affective computing is the study and development of systems and devices that can recognize, interpret, process, and simulate human affects. Computers and other devices being able to detect and identify how people feel would opens the door to an uncountable number of applications and would imply an improvement in the quality of life for society.

In this project the emotion recognition through the electroencephalogram has been studied. Reviews of the state of the art in this subject and machine learning techniques have been carried out to later address the problems of the extraction of characteristics from encephalogram signals and the classification of emotions. The latter has been divided into two problems of binary classification: on the one hand it is classified if the emotion is positive or negative (valence classification) and on the other hand if the emotion implies high or low excitation (arousal classification).

• Dataset used: DEAPdataset.

• Filtering and characteristic extraction from the signals done with Matlab.

• Classification done with Python using Scikit-learn and Keras libraries.

Francisco Javier Ortiz Bonilla - Pogorelich