Stony Brook University Libraries’ AI Research Scientist, Hojun Son Ph.D recently published the article, Assessment of a Multimodal Feedback System for Independent Navigation of People with Visual Impairments in the Journal of Visual Impairments & Blindness.
Son led the study that explores the possibility of wearable devices that could offer faster and more intuitive guidance for people with visual impairments.
Existing navigation systems for people with visual impairments have largely concentrated on destination and object recognition, while overlooking how information can be communicated in ways that align with users’ cognition. Rather than relying solely on verbal feedback, the most widely used approach, this research aims to explore alternative ways to support future navigation systems through the use of smart glasses.
For this study, the system used QR code-based guidance signals in images to generate guidance cues during indoor travel. We evaluated two guidance cue sets, including four multimodal feedback types (verbal + vibration), and four media types on the five different indoor routes up to 180 m of travel with at least two sharp turns. The study measured preferred walking speed, the mental effort involved in following navigation instructions, the number of incorrect responses, and how often assistant intervention was required when users became stuck while navigating the system.
Findings included:
- Head vibration provided instant responses and was more reliable than verbal guidance, likely because of its spatial cues.
- Participants’ preferences did not match their performance. 3 out of 10 preferred verbal feedback, and half preferred belt vibration among the vibration-based options, despite performing better with head vibration.
- The combined feedback system (verbal and vibration) was the most preferred option, chosen by 6 out of 10 participants. Verbal feedback alone was preferred by 3 participants, while 1 participant selected vibration feedback only.
Building on this research, future initiatives can advance library accessibility while providing students with meaningful opportunities to contribute through the Libraries’ soon-to-launched Codex+AI Lab.