A Perception-Action-Semantics Loop for Enhancing Accessible Mixed Reality Heritage Exploration for Blind and Low Vision Users

Cultural heritage exploration for blind and low vision (BLV) users requires coordinated multimodal support rather than isolated sensory substitution. We developed an accessible mixed reality (MR) system integrating tactile exploration, navigation support, and an LLM-based conversational agent, and conducted a counterbalanced mixed-methods study with 20 BLV participants across three interaction modes. Tactile-grounded semantic interaction was associated with stronger aesthetic engagement, greater perceived autonomy, higher flow, and longer artifact-centered exploration without increasing cognitive load. These findings suggest that interaction coherence, rather than modality accumulation, better explains accessible heritage exploration in this site-specific setting, subject to the constraints of current head-mounted MR systems.