Indirect Time of Flight Near Field LiDAR Depth Correction Using Spiking Neural Networks

  • Mena Nagiub
  • , Thorsten Beuth
  • , Ganesh Sistu
  • , Heinrich Gotzig
  • , Ciaran Eising

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Spiking Neural Networks (SNN) is a machine learning model inspired by the spiking nature of biological human brain neurons. These neural models lead to the creation of neuromorphic computing chips, which can execute at a very low power profile, less than 1 Watt. Such a low profile can be very useful in developing low-power sensors. In this paper, we investigate the feasibility of implementing our iToF LiDAR depth correction model for static scenes using SNN, leveraging the advantages of neuromorphic chips to develop low-power sensors suitable for electric vehicles and battery-powered autonomous mobile robots. We present our results, findings, and recommendations to implement such a system. We also discuss the challenges we have faced and the possible solutions to overcome them. The paper has benchmarked the new SNN-based model versus the original ANN model in terms of accuracy, and it has been found that SNN-based models can provide comparable accuracy under certain conditions.

Original languageEnglish
Title of host publication5th International Mobile, Intelligent, and Ubiquitous Computing Conference, MIUCC 2025
EditorsAyman M. Bahaa-Eldin, Ashraf AbdelRaouf, Nada Shorim, Nada Nofal, Yasmine Kandil
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages450-457
Number of pages8
ISBN (Electronic)9798331539221
DOIs
Publication statusPublished - 2025
Event5th International Mobile, Intelligent, and Ubiquitous Computing Conference, MIUCC 2025 - Cairo, Egypt
Duration: 17 Sep 202518 Sep 2025

Publication series

Name5th International Mobile, Intelligent, and Ubiquitous Computing Conference, MIUCC 2025

Conference

Conference5th International Mobile, Intelligent, and Ubiquitous Computing Conference, MIUCC 2025
Country/TerritoryEgypt
CityCairo
Period17/09/2518/09/25

Keywords

  • depth correction
  • LiDAR
  • lowpower
  • neural networks
  • neuromorphic
  • spiking

Fingerprint

Dive into the research topics of 'Indirect Time of Flight Near Field LiDAR Depth Correction Using Spiking Neural Networks'. Together they form a unique fingerprint.

Cite this