Investigating the Non-Verbal Behavior Features of Bullying for the Development of an Automatic Recognition System in Social Virtual Reality

EasyChair Preprint 8108, version 3

Abstract

We investigate the possibilities of automatically detecting social discomfort and social anxiety via non-verbal behaviours in social Virtual Reality (VR). This is important because a well-developed automatic recognition system could facilitate interventions and moderation in social VR without requiring real-time parental supervision. To initially explore this question of recognition, we prototyped a small set of 3D stimuli representing a bullying scenario and explored in a small formative preliminary study what human observers perceived from the stimuli. Future work is required with different problematic situations in social VR and evaluations with more participants before developing an automatic recognition system.

Keyphrases: Social Virtual Reality, automatic analyses of human behavior, bullying, non-verbal behaviors

@booklet{EasyChair:8108,
  author    = {Cristina Fiani and Stacy Marsella},
  title     = {Investigating the Non-Verbal Behavior Features of Bullying for the Development of an Automatic Recognition System in Social Virtual Reality},
  howpublished = {EasyChair Preprint 8108},
  year      = {EasyChair, 2022}}