VIVR: Presence of immersive interaction for visual impairment virtual reality

Abstract

The immersive virtual reality (VR) to provide a realistic walking experience for the visually impaired is proposed in this study. To achieve this, a novel immersive interaction using a walking aid, i.e., a white cane, is designed. The key structure of the proposed interaction consists of a walking process that enables users with visual impairments to process the ground recognition and inference processes realistically by connecting the white cane to the VR controller. Additionally, a decision-making model using deep learning is proposed to design interactions that can be applied to real-life situations instead of being limited to virtual environment experiences. A learning model is designed that can accurately and efficiently process sensing of braille block, which is an important process in the walking of visually impaired people using a white cane assistance tool. The goal is to implement a white cane walking system that can be used in the real world in addition to a virtual environment. Finally, a survey is conducted to confirm that the proposed immersive interaction provides a walking experience with high presence in virtual reality when compared with the real-world experience. The applicability of the proposed deep-learning-based decisionmaking model in the real world is verified by its high accuracy in recognition of braille block.