Context-Aware Navigation Assistant for the Visually Impaired

This paper presented an offline vision-based navigation assistance system designed to improve the independent mobility of visually impaired individuals through real-time environmental perception and intelligent audio guidance. The proposed framework integrates a general object detection model with a specialized staircase detection model to accurately identify navigation-relevant obstacles. A context-aware object prioritization strategy was introduced to filter unnecessary detections and communicate only the nearest and most significant obstacles, thereby reducing cognitive overload and improving the effectiveness of audio feedback. Furthermore, the system estimates the relative position of detected objects and generates concise Hindi voice instructions to assist users in navigating safely within their surroundings. Since the entire framework operates offline, it eliminates dependence on internet connectivity while ensuring reliable and low-latency performance. The proposed system provides a practical and scalable solution for real-time assistive navigation. Future work will focus on improving distance estimation accuracy, integrating depth-sensing techniques, enhancing dynamic obstacle recognition, and expanding multilingual voice support to further improve navigation efficiency and user experience.

[1]     M. Yang, M. Dong, C. Li, and F. Nie, “YOLO-RDFEA: Object detection in RD imagery with improved YOLOv8 based on feature enhancement and attention mechanisms,” IEEE Access, vol. 12, pp. 158226–158238, 2024, doi: 10.1109/ACCESS.2024.3485499. [2]     G. Yang and J. Saniie, “Sight-to-sound human-machine interface for guiding and navigating visually impaired people,” IEEE Access, vol. 8, pp. 26612–26622, 2020, doi: 10.1109/ACCESS.2020.2970979. [3]     U. Masud, T. Saeed, H. M. Malaikah, F. ul Islam, and G. Abbas, “Smart assistive system for visually impaired people: Obstruction avoidance through object detection and classification,” IEEE Access, vol. 10, pp. 13428–13440, 2022, doi: 10.1109/ACCESS.2022.3146320. [4]     X. Leong and R. K. Ramasamy, “Obstacle detection and distance estimation for visually impaired people,” IEEE Access, vol. 11, pp. 136609–136629, 2023, doi: 10.1109/ACCESS.2023.3338154. [5]     S. Ikram, I. S. Bajwa, S. Gyawali, A. Ikram, and N. Alsubaie, “Multimodality-based situational knowledge for obstacle detection and alert generation to enhance navigation assistive systems,” IEEE Access, vol. 13, pp. 1–15, 2025, doi: 10.1109/ACCESS.2025.3367421. [6]     R. Wazirali and E. Alkhamash, “Enhancing object detection in assistive technology for the visually impaired: A DETR-based approach,” IEEE Access, vol. 4, pp. 1–10, 2016, doi: 10.1109/ACCESS.2016.2586308. [7]     F. Ashiq, M. Asif, M. Bin Ahmad, S. Zafar, K. Masood, T. Mahmood, M. T. Mahmood, and I. H. Lee, “CNN-based object recognition and tracking system to assist visually impaired people,” IEEE Access, vol. 10, pp. 14819–14834, 2022, doi: 10.1109/ACCESS.2022.3148036. [8]     A. Bahrami, M. Yazdani, and A. Fotouhi, “Efficient real-time pathfinding for visually impaired individuals,” IEEE Access, vol. 13, pp. 1–12, 2025, doi: 10.1109/ACCESS.2025.3351987. [9]     P. K. Sahu and R. Tiwari, “Assistive devices analysis for visually impaired persons: A review,” Int. J. Sci. Technol. Res., vol. 9, 2021. [10]   W. M. Elmannai and K. M. Elleithy, “A highly accurate and reliable data fusion framework for guiding the visually impaired,” IEEE Access, vol. 13, pp. 1–15, 2025, doi: 10.1109/ACCESS.2025.3365129.