A Critical Review of Deep Learning-Based Multi-Sensor Fusion Techniques

Benedict Marsh; Abdul Sadka; Hamid Bahai

doi:10.3390/s22239364

A Critical Review of Deep Learning-Based Multi-Sensor Fusion Techniques

Benedict Marsh, Abdul Sadka, Hamid Bahai

College of Engineering and Physical Sciences

Research output: Contribution to journal › Article › peer-review

Abstract

In this review, we provide a detailed coverage of multi-sensor fusion techniques that use RGB stereo images and a sparse LiDAR-projected depth map as input data to output a dense depth map prediction. We cover state-of-the-art fusion techniques which, in recent years, have been deep learning-based methods that are end-to-end trainable. We then conduct a comparative evaluation of the state-of-the-art techniques and provide a detailed analysis of their strengths and limitations as well as the applications they are best suited for.

Original language	English
Article number	9364
Journal	Sensors
Volume	22
Issue number	23
DOIs	https://doi.org/10.3390/s22239364
Publication status	Published - 1 Dec 2022

Bibliographical note

© 2022 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).

Access to Document

10.3390/s22239364Licence: CC BY 4.0

A Critical Review of Deep Learning-Based Multi-Sensor Fusion Techniques
© 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).
Final published version, 9.56 MBLicence: CC BY 4.0

Cite this

@article{0e87b17865eb4cd7ada78294cb0aab0f,

title = "A Critical Review of Deep Learning-Based Multi-Sensor Fusion Techniques",

abstract = "In this review, we provide a detailed coverage of multi-sensor fusion techniques that use RGB stereo images and a sparse LiDAR-projected depth map as input data to output a dense depth map prediction. We cover state-of-the-art fusion techniques which, in recent years, have been deep learning-based methods that are end-to-end trainable. We then conduct a comparative evaluation of the state-of-the-art techniques and provide a detailed analysis of their strengths and limitations as well as the applications they are best suited for.",

author = "Benedict Marsh and Abdul Sadka and Hamid Bahai",

note = "{\textcopyright} 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/). ",

year = "2022",

month = dec,

day = "1",

doi = "10.3390/s22239364",

language = "English",

volume = "22",

journal = "Sensors",

issn = "1424-8239",

publisher = "MDPI AG",

number = "23",

}

TY - JOUR

T1 - A Critical Review of Deep Learning-Based Multi-Sensor Fusion Techniques

AU - Marsh, Benedict

AU - Sadka, Abdul

AU - Bahai, Hamid

N1 - © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

PY - 2022/12/1

Y1 - 2022/12/1

N2 - In this review, we provide a detailed coverage of multi-sensor fusion techniques that use RGB stereo images and a sparse LiDAR-projected depth map as input data to output a dense depth map prediction. We cover state-of-the-art fusion techniques which, in recent years, have been deep learning-based methods that are end-to-end trainable. We then conduct a comparative evaluation of the state-of-the-art techniques and provide a detailed analysis of their strengths and limitations as well as the applications they are best suited for.

AB - In this review, we provide a detailed coverage of multi-sensor fusion techniques that use RGB stereo images and a sparse LiDAR-projected depth map as input data to output a dense depth map prediction. We cover state-of-the-art fusion techniques which, in recent years, have been deep learning-based methods that are end-to-end trainable. We then conduct a comparative evaluation of the state-of-the-art techniques and provide a detailed analysis of their strengths and limitations as well as the applications they are best suited for.

UR - https://www.mdpi.com/1424-8220/22/23/9364

U2 - 10.3390/s22239364

DO - 10.3390/s22239364

M3 - Article

SN - 1424-8239

VL - 22

JO - Sensors

JF - Sensors

IS - 23

M1 - 9364

ER -

A Critical Review of Deep Learning-Based Multi-Sensor Fusion Techniques

Abstract

Bibliographical note

Access to Document

Other files and links

Fingerprint

Cite this