Abstract
Facial tracking has been an active subject of research with applications in security systems, pain monitoring, human-robot interactions, character animation, and posttraumatic stress disorder amongst others. Furthermore, facial expressions allow communicating emotions that are relevant to social interactions. In particular, children with autism can have difficulties interpreting facial expressions that can affect their communication. Currently, approaches to improve the autistic child's recognition of emotions include therapy that requires them to identify faces based on pictures embedded in computer and mobile applications, and the assessment is subjective corresponding to the observations reported by the subject. In this paper, we present a preliminary study that explores the use of braincomputer interfaces to detect happy, sad, angry, and surprise facial expressions. The preliminary results indicate that brain signals can be used to monitor face expression changes under certain conditions, although special care has to be considered when processing the data given the large amount of noise caused by muscle movement and multimodal stimuli.