Effects of lidar and radar resolution on DNN-based vehicle detection

Itai Orr; Harel Damari; Meir Halachmi; Mark Raifel; Kfir Twizer; Moshik Cohen; Zeev Zalevsky

doi:https://doi.org/10.1364/josaa.431582

Effects of lidar and radar resolution on DNN-based vehicle detection

Itai Orr, Harel Damari, Meir Halachmi, Mark Raifel, Kfir Twizer, Moshik Cohen, Zeev Zalevsky

Bar-Ilan University, Technion - Israel Institute of Technology

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

Abstract

Vehicle detection plays a critical role in autonomous driving, where two central sensing modalities are lidar and radar. Although many deep neural network (DNN)-based methods have been proposed to solve this task, a systematic and methodological examination on the influence of the data on those methods is still missing. In this work, we examine the effects of resolution on the performance of vehicle detection for both lidar and radar sensors. We propose subsampling methods that can improve the performance and efficiency of DNN-based solutions and offer an alternative approach to traditional sensor-design trade-offs.

Original language	English
Pages (from-to)	B29-B36
Journal	Journal of the Optical Society of America A: Optics and Image Science, and Vision
Volume	38
Issue number	10
DOIs	https://doi.org/10.1364/josaa.431582
State	Published - 1 Oct 2021

All Science Journal Classification (ASJC) codes

Electronic, Optical and Magnetic Materials
Atomic and Molecular Physics, and Optics
Computer Vision and Pattern Recognition

Access to Document

https://doi.org/10.1364/josaa.431582

Cite this

@article{c16275c8267a49f08850294f4f39f731,

title = "Effects of lidar and radar resolution on DNN-based vehicle detection",

abstract = "Vehicle detection plays a critical role in autonomous driving, where two central sensing modalities are lidar and radar. Although many deep neural network (DNN)-based methods have been proposed to solve this task, a systematic and methodological examination on the influence of the data on those methods is still missing. In this work, we examine the effects of resolution on the performance of vehicle detection for both lidar and radar sensors. We propose subsampling methods that can improve the performance and efficiency of DNN-based solutions and offer an alternative approach to traditional sensor-design trade-offs.",

author = "Itai Orr and Harel Damari and Meir Halachmi and Mark Raifel and Kfir Twizer and Moshik Cohen and Zeev Zalevsky",

note = "Publisher Copyright: {\textcopyright} 2021 Optica Publishing Group.",

year = "2021",

month = oct,

day = "1",

doi = "https://doi.org/10.1364/josaa.431582",

language = "الإنجليزيّة",

volume = "38",

pages = "B29--B36",

journal = "Journal of the Optical Society of America A: Optics and Image Science, and Vision",

issn = "1084-7529",

publisher = "The Optical Society",

number = "10",

}

TY - JOUR

T1 - Effects of lidar and radar resolution on DNN-based vehicle detection

AU - Orr, Itai

AU - Damari, Harel

AU - Halachmi, Meir

AU - Raifel, Mark

AU - Twizer, Kfir

AU - Cohen, Moshik

AU - Zalevsky, Zeev

PY - 2021/10/1

Y1 - 2021/10/1

N2 - Vehicle detection plays a critical role in autonomous driving, where two central sensing modalities are lidar and radar. Although many deep neural network (DNN)-based methods have been proposed to solve this task, a systematic and methodological examination on the influence of the data on those methods is still missing. In this work, we examine the effects of resolution on the performance of vehicle detection for both lidar and radar sensors. We propose subsampling methods that can improve the performance and efficiency of DNN-based solutions and offer an alternative approach to traditional sensor-design trade-offs.

AB - Vehicle detection plays a critical role in autonomous driving, where two central sensing modalities are lidar and radar. Although many deep neural network (DNN)-based methods have been proposed to solve this task, a systematic and methodological examination on the influence of the data on those methods is still missing. In this work, we examine the effects of resolution on the performance of vehicle detection for both lidar and radar sensors. We propose subsampling methods that can improve the performance and efficiency of DNN-based solutions and offer an alternative approach to traditional sensor-design trade-offs.

UR - http://www.scopus.com/inward/record.url?scp=85116396977&partnerID=8YFLogxK

U2 - https://doi.org/10.1364/josaa.431582

DO - https://doi.org/10.1364/josaa.431582

M3 - مقالة

C2 - 34612987

SN - 1084-7529

VL - 38

SP - B29-B36

JO - Journal of the Optical Society of America A: Optics and Image Science, and Vision

JF - Journal of the Optical Society of America A: Optics and Image Science, and Vision

IS - 10

ER -

Effects of lidar and radar resolution on DNN-based vehicle detection

Abstract

All Science Journal Classification (ASJC) codes

Access to Document

Other files and links

Fingerprint

Cite this