Researchers from the Computer Science Centre at the University of Basel in Switzerland have unveiled an artificial intelligence system designed to identify basic human emotions based on facial expressions, which detected expressions that had gone unnoticed by doctors.
The team developed an AI system capable of recognizing six fundamental emotions: happiness, surprise, anger, disgust, sadness, and fear, utilizing over 30,000 images of facial expressions.
This system was tested by analyzing video clips from psychotherapy sessions of 23 patients suffering from psychological disorders.
Martin Stepan, a psychologist at the university, stated that the new system was fed more than 950 hours of video recordings involving psychiatric patients. The results were astonishing, as the system's analysis concurred with conclusions drawn by three physicians.
The system also managed to capture certain facial expressions that appeared for mere fractions of a second, such as a fleeting smile or a gesture of disgust, which the doctors had overlooked.
Researchers believe that this new artificial intelligence system has succeeded in detecting facial expressions as a reliable means of gauging emotions and thus could serve as an important tool for research and treatment.
Facial expressions are known to reflect a person's emotional state and are considered a critical element of psychotherapy for understanding an individual's feelings at a given moment.
