Gaze-based Anxiety Sensitive Virtual Social Communication Platform for Individuals with Autism
Source
Conference on Human Factors in Computing Systems Proceedings
Date Issued
2022-04-28
Author(s)
Krishnappa Babu, Pradeep Raj
Abstract
Difficulties in social communication along with impairments in understanding other's emotions and exhibiting atypical gaze behavior while avoiding social cues are common concerns in individuals with autism spectrum disorder (ASD). Such deficits are reported to be related with social anxiety. Social anxiety often deters the process of skill learning. Given the dramatic rise in the prevalence of ASD and limited availability of trained therapists, getting access to technology-assisted platforms that can estimate one's anxiety and autonomously vary the task challenges for effective skill learning is critical. The potential of gaze-related indices to serve as biomarkers to one's anxiety, and advancement in computing technologies allowing real-time access to gaze-related social signals, we have designed a Virtual Reality (VR)-based anxiety-sensitive social communication task platform. This features a Rule Generator that is used to offer tasks based on the composite effect of task performance and estimated anxiety from one's gaze in an individualized manner. Results of a study with 10 individuals with ASD showed that the anxiety-sensitive system showed promise in eliciting improved task performance and increased looking towards the face region of a communicator displaying emotional expression in comparison to the currently existing system that is blind to one's anxiety. This paves the way towards design of a complementary tool for the therapists.
Subjects
Affective Computing | Anxiety | Autism | Flow Theory | Looking Pattern | Virtual Reality
