Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

Federated Learning for Obstacle Detection to Assist the Visually-Impaired Using Augmented Reality

Akbarian, Fatemeh LU ; Vlaeminck, Robbe ; Verheijen, Joran ; Aminifar, Amin and Aminifar, Amir LU orcid (2025)
Abstract
Vision impairment increases risks such as social isolation, mobility challenges, and falls. Wearable Augmented Reality (AR) devices with Artificial Intelligence (AI) can enhance sensory perception by enabling real-time recognition of obstacles, assisting visually impaired individuals during street navigation, aiming to reduce the risk of falls. In this paper, we propose a Federated Learning (FL) framework for outdoor obstacle detection using resource-constrained edge AR devices. As such, our proposed framework is designed to optimize energy efficiency by partially fine-tuning a generic pre-trained model originally developed for visually impaired assistance. Our framework is evaluated on a testbed with NVIDIA Jetson Nano, demonstrating that... (More)
Vision impairment increases risks such as social isolation, mobility challenges, and falls. Wearable Augmented Reality (AR) devices with Artificial Intelligence (AI) can enhance sensory perception by enabling real-time recognition of obstacles, assisting visually impaired individuals during street navigation, aiming to reduce the risk of falls. In this paper, we propose a Federated Learning (FL) framework for outdoor obstacle detection using resource-constrained edge AR devices. As such, our proposed framework is designed to optimize energy efficiency by partially fine-tuning a generic pre-trained model originally developed for visually impaired assistance. Our framework is evaluated on a testbed with NVIDIA Jetson Nano, demonstrating that FL improves accuracy over standalone models on each AR device, while achieving an accuracy comparable to centralized approaches, but without the need to transfer the local raw data on each AR device to a central server/cloud to alleviate privacy concerns. (Less)
Please use this url to cite or link to this publication:
author
; ; ; and
organization
publishing date
type
Contribution to conference
publication status
in press
subject
keywords
Federated learning (FL), Obstacle detection, Assistive technology (AT), Vision impairment, Augmented Reality (AR), Virtual reality (VR)
language
English
LU publication?
yes
id
1afe41a8-2f57-4f3b-8d20-bf4d359f865b
date added to LUP
2025-12-10 14:06:53
date last changed
2025-12-15 10:26:20
@misc{1afe41a8-2f57-4f3b-8d20-bf4d359f865b,
  abstract     = {{Vision impairment increases risks such as social isolation, mobility challenges, and falls. Wearable Augmented Reality (AR) devices with Artificial Intelligence (AI) can enhance sensory perception by enabling real-time recognition of obstacles, assisting visually impaired individuals during street navigation, aiming to reduce the risk of falls. In this paper, we propose a Federated Learning (FL) framework for outdoor obstacle detection using resource-constrained edge AR devices. As such, our proposed framework is designed to optimize energy efficiency by partially fine-tuning a generic pre-trained model originally developed for visually impaired assistance. Our framework is evaluated on a testbed with NVIDIA Jetson Nano, demonstrating that FL improves accuracy over standalone models on each AR device, while achieving an accuracy comparable to centralized approaches, but without the need to transfer the local raw data on each AR device to a central server/cloud to alleviate privacy concerns.}},
  author       = {{Akbarian, Fatemeh and Vlaeminck, Robbe and Verheijen, Joran and Aminifar, Amin and Aminifar, Amir}},
  keywords     = {{Federated learning (FL); Obstacle detection; Assistive technology (AT); Vision impairment; Augmented Reality (AR); Virtual reality (VR)}},
  language     = {{eng}},
  month        = {{07}},
  title        = {{Federated Learning for Obstacle Detection to Assist the Visually-Impaired Using Augmented Reality}},
  url          = {{https://lup.lub.lu.se/search/files/235900236/Federated_Learning_for_Obstacle_Detection_to_Assist_the_Visually-Impaired_Using_Augmented_Reality.pdf}},
  year         = {{2025}},
}