Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Real-Time Assistive Navigation for the Visually Impaired: A Scalable Approach for Indoor and Outdoor Mobility

Dabbrata Das, Argho Deb Das, Farhan Sadaf, Azhar Uddin, Tirtho Mondal

TL;DR

This work tackles real-time indoor/outdoor navigation for blind and low-vision users by proposing PathFinder, a mapless, fully offline smartphone system that uses monocular depth estimation and a novel depth-based pathfinding algorithm to identify the longest obstacle-free route with low computational cost. It benchmarks six AI-assisted approaches against the proposed method, showing PathFinder achieves favorable trade-offs between accuracy and response time, and remains scalable without internet access. A usability study with 15 BLV participants reveals strong acceptance, with 73% learning the app in under a minute and 80% praising its accuracy and responsiveness, highlighting practical impact and independence from cloud services. Overall, PathFinder offers a low-cost, reliable alternative to cloud-dependent navigation tools, with demonstrated potential to reduce barriers to independent mobility in both indoor and outdoor environments, while noting the need to address low-light and highly dynamic scenarios in future work.

Abstract

Navigating unfamiliar environments remains one of the most persistent and critical challenges for people who are blind or have limited vision (BLV). Existing assistive tools often rely on online services or APIs, making them costly, internet-dependent, and less reliable in real-time use. To address these limitations, we propose PathFinder, a novel mapless mobile phone-based navigation system that operates fully offline. Our method processes monocular depth images and applies an efficient pathfinding algorithm to identify the longest, clearest obstacle-free route, ensuring optimal navigation with low computational cost. Comparative evaluations show that PathFinder reduces mean absolute error (MAE), speeds decision-making, and achieves real-time responsiveness indoors and outdoors. A usability study with 15 BLV participants confirmed its practicality, where 73% learned to operate it in under a minute, and 80% praised its accuracy, responsiveness, and convenience. Despite challenges in complex indoor layouts and low light, PathFinder offers a low-cost, scalable, reliable alternative.

Real-Time Assistive Navigation for the Visually Impaired: A Scalable Approach for Indoor and Outdoor Mobility

TL;DR

This work tackles real-time indoor/outdoor navigation for blind and low-vision users by proposing PathFinder, a mapless, fully offline smartphone system that uses monocular depth estimation and a novel depth-based pathfinding algorithm to identify the longest obstacle-free route with low computational cost. It benchmarks six AI-assisted approaches against the proposed method, showing PathFinder achieves favorable trade-offs between accuracy and response time, and remains scalable without internet access. A usability study with 15 BLV participants reveals strong acceptance, with 73% learning the app in under a minute and 80% praising its accuracy and responsiveness, highlighting practical impact and independence from cloud services. Overall, PathFinder offers a low-cost, reliable alternative to cloud-dependent navigation tools, with demonstrated potential to reduce barriers to independent mobility in both indoor and outdoor environments, while noting the need to address low-light and highly dynamic scenarios in future work.

Abstract

Navigating unfamiliar environments remains one of the most persistent and critical challenges for people who are blind or have limited vision (BLV). Existing assistive tools often rely on online services or APIs, making them costly, internet-dependent, and less reliable in real-time use. To address these limitations, we propose PathFinder, a novel mapless mobile phone-based navigation system that operates fully offline. Our method processes monocular depth images and applies an efficient pathfinding algorithm to identify the longest, clearest obstacle-free route, ensuring optimal navigation with low computational cost. Comparative evaluations show that PathFinder reduces mean absolute error (MAE), speeds decision-making, and achieves real-time responsiveness indoors and outdoors. A usability study with 15 BLV participants confirmed its practicality, where 73% learned to operate it in under a minute, and 80% praised its accuracy, responsiveness, and convenience. Despite challenges in complex indoor layouts and low light, PathFinder offers a low-cost, scalable, reliable alternative.
Paper Structure (48 sections, 3 equations, 11 figures, 9 tables)

This paper contains 48 sections, 3 equations, 11 figures, 9 tables.

Table of Contents

  1. Introduction
  2. Related Works
  3. Navigation Systems for BLV Individuals
  4. AI-Powered Scene Description and Accessibility Tools
  5. User Engagement and Broader Assistive Technologies
  6. Gaps in Existing Research
  7. Methodology
  8. Navigation Schemes
  9. Left/Forward/Right (DOF 3)
  10. In Degree (DOF 180)
  11. Clock Hour Hand (DOF 13)
  12. AI-Assisted Navigation Approaches
  13. Approach 1: 2D RGB Image to Vision Language Models (VLMs)
  14. Approach 2: Depth Images to VLMs
  15. Approach 3: Combined RGB and Depth Images to VLMs
  16. ...and 33 more sections

Figures (11)

  • Figure 1: Visualization of three navigation schemes to define navigation direction.
  • Figure 2: Block diagram of all our approaches for identifying the best path for BLV navigation.
  • Figure 3: Direction annotation using a clock-hand overlay, where the nearest free space to the virtual hour hand is labeled as the ground-truth navigation direction.
  • Figure 4: End-to-end annotation workflow used to generate the dataset labels.
  • Figure 5: Process for optimal pathfinding calculation illustrating path exploration and optimal route selection.
  • ...and 6 more figures