Abstract
People with visual impairments (PVI) increasingly rely on camera-enabled smartphone apps for tasks like photography, navigation, and text recognition. Despite the growing use of these applications, precise camera aiming remains a significant challenge. This study explores the impact of virtual reality (VR) exploration compared to traditional text/audio (TA) instructions in the context of learning to use a walk-light detector app at traffic intersections. We developed a VR exploration tool based on insights gathered from interviews with PVI. A user study was conducted, involving 13 PVI participants divided into two groups: VR exploration and TA instructions. Following indoor training using the respective approaches, participants from both groups used the walk light detector app outdoors. According to the participants' subjective feedback, a higher proportion of participants in the TA group found the training easier, potentially due to shortcomings in our VR protocol and differences between the real world and VR. However, more VR participants gained insights into walk light detection and felt unable to use the detector without VR training, compared to the TA group.