Semantic and Spatial Sound-Object Recognition for Assistive Navigation

This project presents an interactive repository designed to explore how sound can assist in navigating urban and indoor spaces, particularly for blind or visually impaired (BVI) individuals. Building on spatial cognition and auditory perception, the sonification method maps object properties, like material, width, interaction, distance, and position, to specific sound cues using pitch, timbre, gain, and panning. This web tool allows users to explore object-sonification pairings used during the experimental phases of the research.

The method was validated through a three-phase study: sound selection using acoustic feature analysis, an online perceptual survey on sound-object associations, and virtual reality-based navigation tests. Results showed strong differentiation in material perception and promising navigation accuracy, though they also highlighted areas for further refinement. The project aims to inform future assistive technologies by enriching spatial awareness through sound-based cues.