Lune

UbiComp2026顶会

VR360-SSL: Towards Accessible and Clinically-Viable 360° Sound Source Localization in Virtual Reality

Yihan Yin, Zhimin Wang, Ying Hu, Ke Zhang, Feng Lu

2026年份

摘要

Sound source localization (SSL) capability serves as a crucial indicator of early hearing impairment, with applications in hearing disorder diagnosis, hearing aid fitting, and neurodegenerative disease screening. However, traditional SSL testing requires expensive professional equipment (over $500,000) and dedicated soundproof facilities (over 30 m 2 ), significantly limiting accessibility for large-scale screening programs. While VR-based systems offer a promising alternative, they face fundamental challenges when extending to full 360° scenarios, particularly severe confusion between front and rear sound sources and interaction fatigue for rear target selection. To address these limitations, we propose the VR360-SSL system, the first 360° VR-based screening and rehabilitation tracking tool. Our approach introduces a frequency domain enhancement strategy for non-individualized head-related transfer functions that perceptually compensates for front-back ambiguity by reinforcing high-frequency cues for front sources and enhancing low-frequency components for rear sources. We further incorporate adaptive interaction modalities (eye gaze and hand gesture) with pre-experiment assessment that matches interaction modes to users' VR experience and spatial imagination abilities. Preliminary experiments with 32 participants (including 23 normal hearing (NH) individuals and 9 single-sided deafness (SSD) patients) indicate that the system provides localization trends consistent with real sound field environments and successfully identifies performance gradients across different hearing impairment severities. Front-back confusion rates were significantly reduced to levels comparable with real-world medical setups. By reducing costs to a single VR headset, our system enables SSL testing expansion from specialized clinics to primary care and community-based screening scenarios, particularly suitable for VR-naive elderly populations and resource-limited areas.

问问这篇 Paper

问问你的智能体。

Lune 读过与它相关的顶会 Paper,每个回答都会注明依据哪几篇。

可以从这些问题问起

智能体调用

Lunesearch_papers

在 Lune 里问

免费开始,无需绑卡

相关 Paper

黄昏的海面,两侧是细线勾勒的悬崖