Looking to the Future through Facial Emotion Recognition

This dissertation provides an analysis of an emerging field of Artificial Intelligence, namely Facial Emotion Recognition (hereinafter “FER”). In a nutshell, this cutting-edge technology is able to identify how a person is feeling just by capturing and evaluating her facial expressions. Firstly, an analysis of the concept of FER and its functioning is provided. The reliability of FER systems is examined, especially considering Paul Ekman's Basic Emotions Theory. Concerns and perils associated with FER systems are then explored, encompassing feasibility and biases, data accuracy, manipulation, privacy, transparency, and control. The dissertation investigates the various fields of application where FER systems are employed, such as the personalized recommendations in domains like music selection and cultural site recommendations, education applications for monitoring students' attention, employment applications for decision-making processes, public safety uses such as crime detection and lie detection at border control, and driving fatigue detection. Moreover, one important aspect addressed in this dissertation is the discrimination as potential counter-effect that can arise from the usage of such systems. Accordingly, an analysis of racial discrimination, discrimination against people with disabilities, and gender discrimination is provided. Data protection and privacy concerns are key considerations throughout the dissertation. A comprehensive overview of the right to data protection and the right to privacy is provided, with a specific focus on the European Union's General Data Protection Regulation and the proposed Artificial Intelligence Act with its risk-based approach. This dissertation examines the uses, difficulties, and ethical considerations of Facial Emotion Recognition systems in an effort to provide a thorough grasp of the current situation and potential future repercussions of this ground-breaking technology. By performing critical analysis and review, it may be possible to manage the complexity around FER systems, aiming to maximize their benefits while minimizing potential risks and guaranteeing an ethical and inclusive future.