NeuroLens: A Cognitive-aware Assistive Navigation Framework for the Visually Impaired using EEG and 3D Spatial Perception

Abstract

Globally, over 250 million people live with moderate to severe visual impairments, facing significant barriers to safe and independent navigation. While assistive tools such as canes, guide dogs, and smartphone-based apps exist, these solutions often require physical interaction, lack environmental awareness, and fail to adapt to user intent or cognitive state. This conceptual research proposes a high-level architecture for "NeuroLens," a brain-computer interface (BCI)-driven navigation framework that integrates noninvasive EEG signals with real-time 3D spatial perception using depth cameras. The goal is to enhance hands-free mobility for visually impaired users by inferring navigation intent and delivering contextual feedback via audio and haptic channels. Unlike existing AR or BCI-based tools, NeuroLens introduces a dual-input decision mechanism and continuous real-time deviation monitoring through an FSM controller. Although no prototype has been implemented, the system design is grounded in prior literature and optimized for mobile platforms. Future work will focus on prototype development, performance benchmarking, and real-world usability evaluation to validate the framework’s feasibility and effectiveness.