Dr Eng. Teresa Zawadzka from the Faculty of Electronics, Telecommunications and Informatics at the Gdansk University of Technology is conducting research with her students into the possibilities of reading emotions using graph neural networks. Because of their extraordinary expressive capabilities, graph neural networks are extremely helpful in the field of machine learning.

Graph structures allow the storage of a great variety of data and the relationships between them – not only the emotions, but also the context of their occurrence. By context here, we mean both the person to whom the emotion relates and the circumstances in which the emotion arose.

“The needs and applications for AI machines are numerous. Let’s start with sick people, e.g. in a coma or after a stroke, who cannot express by facial expression or verbally that they feel pain or anxiety. A proper analysis of certain monitoring parameters of their health can show what emotions they are dealing with”, says the researcher.

Read more: https://pg.edu.pl/aktualnosci/2022-09/technologie-i-emocje-badania-na-grafowych-sieciach-neuronowych