Social virtual reality often relies on visual avatar behaviors such as head orientation, spatial arrangement, and turn-taking cues, creating barriers for blind and low vision users in small-group interaction. This project explores private non-visual addressivity cues that help BLV users perceive when attention is directed toward them without publicly marking assistance needs. Through formative co-design with BLV participants and VR designers, we identified breakdowns around speaker attention, turn readiness, and being acknowledged. We then implemented orientation-based cues delivered through either haptic feedback or spatial audio and evaluated them in an exploratory within-subject study with 11 BLV participants. Both cue conditions improved perceived access to conversational attention relative to baseline. Audio cues appeared useful for conversational flow and synchrony, while haptic cues more consistently supported perceived inclusion and social presence. These findings suggest that accessible social VR should treat modality, privacy, density, and user control as core design dimensions.
…
Recent accessibility work has begun to examine how BLV users perceive other people in social VR. VRBubble focuses on proxemic awareness by using audio to convey nearby avatars and their movement [7]. Sighted guidance systems support navigation and contextual interpretation through assistance from another person, while raising questions about agency and scalability [4]. Jung et al. translate several nonverbal behaviors, including eye contact, head nodding, and head shaking, into audio and haptic representations [8]. SocialCueSwitch further emphasizes user-configurable mappings of social cues across sensory modalities [12]. These systems do not necessarily make assistance publicly visible, and we do not characterize them as public cueing systems.
Our distinction instead concerns their primary design focus. Prior work has largely examined proximity awareness, general translation of nonverbal behavior, human guidance, or cross-modal configurability. We examine the narrower interactional problem of privately informing a BLV user that attention is being directed specifically toward them during mixed-ability group conversation. Privacy in this study therefore refers not only to the sensory channel used, but also to reducing reliance on explicit verbal nomination, visible intervention, or assistance initiated by sighted participants.
This project examines private non-visual addressivity cues as one way to support social coordination, addressivity, and participation in mixed-ability social VR conversation. We use private to describe cues that are delivered to the BLV user without requiring other participants to verbally nominate them, explicitly provide assistance, or make the cue visible as a special accommodation in the group interaction. In social settings, such design choices matter because cues can also shape how attention, disability, assistance, and participation are disclosed or noticed by others [9]. Accessibility research likewise emphasizes designing around users’ abilities and preferences rather than forcing users to adapt to a single interaction model [13]. We therefore focus on orientation-based cues that preserve the social function of addressivity, turn readiness, and being noticed in a way that can be perceived non-visually [10].
Whatever you feel about VR I think there is something beautiful about this!