Human Detection in Depth Map Created from Point Cloud

Simple item page
dc.contributor.authorLigocki, Adamcs
dc.contributor.authorŽalud, Luděkcs
dc.date.accessioned2022-02-14T11:55:41Z
dc.date.available2022-02-14T11:55:41Z
dc.date.issued2022-04-02cs
dc.description.abstractThis paper deals with human detection in the LiDAR data using the YOLO object detection neural network architecture. RGB-based object detection is the most studied topic in the field of neural networks and autonomous agents. However, these models are very sensitive to even minor changes in the weather or light conditions if the training data do not cover these situations. This paper proposes to use the LiDAR data as a redundant, and more condition invariant source of object detections around the autonomous agent. We used the publically available real-traffic dataset that simultaneously captures data from RGB camera and 3D LiDAR sensors during the clear-sky day and rainy night time and we aggregate the LiDAR data for a short period to increase the density of the point cloud. Later we projected these point cloud by several projection models, like pinhole camera model, cylindrical projection, and bird-view projection, into the 2D image frame, and we annotated all the images. As the main experiment, we trained the several YOLOv5 neural networks on the data captured during the day and validate the models on the mixed day and night data to study the robustness and information gain during the condition changes of the input data. The results show that the LiDAR-based models provide significantly better performance during the changed weather conditions than the RGB-based models.en
dc.formattextcs
dc.format.extent1-12cs
dc.format.mimetypeapplication/pdfcs
dc.identifier.citationLecture Notes in Computer Science. 2022, p. 1-12.en
dc.identifier.doi10.1007/978-3-030-98260-7_16cs
dc.identifier.isbn9783030982607cs
dc.identifier.issn0302-9743cs
dc.identifier.other172819cs
dc.identifier.urihttp://hdl.handle.net/11012/203906
dc.language.isoencs
dc.relation"European Union (EU)" & "Horizon 2020"en
dc.relation.ispartofLecture Notes in Computer Sciencecs
dc.relation.projectIdinfo:eu-repo/grantAgreement/EC/H2020/857306/EU//RICAIPen
dc.relation.projectIdinfo:eu-repo/grantAgreement/EC/H2020/877539/EU//ArchitectECA2030en
dc.rightsCreative Commons Attribution 4.0 Internationalcs
dc.rights.accessopenAccesscs
dc.rights.sherpahttp://www.sherpa.ac.uk/romeo/issn/0302-9743/cs
dc.rights.urihttp://creativecommons.org/licenses/by/4.0/cs
dc.subjectLiDAR dataen
dc.subjectRGB cameraen
dc.subjectPoint Clouden
dc.subjectprojectionen
dc.subjectYOLOen
dc.subjectOb-ject Detectionen
dc.subjectNeural Networken
dc.subjectDCNNen
dc.titleHuman Detection in Depth Map Created from Point Clouden
dc.type.driverconferenceObjecten
dc.type.statusPeer-revieweden
dc.type.versionsubmittedVersionen
sync.item.dbidVAV-172819en
sync.item.dbtypeVAVen
sync.item.insts2022.06.14 16:54:21en
sync.item.modts2022.06.14 16:14:14en
thesis.grantorVysoké učení technické v Brně. Středoevropský technologický institut VUT. Kybernetika a robotikacs
thesis.grantorVysoké učení technické v Brně. Fakulta elektrotechniky a komunikačních technologií. Ústav automatizace a měřicí technikycs
Files
Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
MESAS2021.pdf
Size:
10.8 MB
Format:
Adobe Portable Document Format
Description:
MESAS2021.pdf
Download
Collections
Ústav automatizace a měřicí techniky
Kybernetika a robotika