POAT-Net: Parallel Offset-Attention Assisted Transformer for 3D Object Detection for Autonomous Driving

3D object detection is playing a key role in the perception process of autonomous driving and industrial robots automation. Inherent characteristics of point cloud raise an enormous challenge to both spatial representation and association analysis. Unordered point cloud spatial data structure and de...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Jinyang Wang, Xiao Lin, Hongying Yu
Formato:	article
Lenguaje:	EN
Publicado:	IEEE 2021
Materias:	3D object detection non-uniform density parallel offset-attention point cloud transformer Electrical engineering. Electronics. Nuclear engineering TK1-9971
Acceso en línea:	https://doaj.org/article/37203ad23a904e6aa2175a9f34c8069d
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:37203ad23a904e6aa2175a9f34c8069d
record_format	dspace
spelling	oai:doaj.org-article:37203ad23a904e6aa2175a9f34c8069d2021-11-17T00:00:26ZPOAT-Net: Parallel Offset-Attention Assisted Transformer for 3D Object Detection for Autonomous Driving2169-353610.1109/ACCESS.2021.3127234https://doaj.org/article/37203ad23a904e6aa2175a9f34c8069d2021-01-01T00:00:00Zhttps://ieeexplore.ieee.org/document/9611257/https://doaj.org/toc/2169-35363D object detection is playing a key role in the perception process of autonomous driving and industrial robots automation. Inherent characteristics of point cloud raise an enormous challenge to both spatial representation and association analysis. Unordered point cloud spatial data structure and density variations caused by gradually varying distances to LiDAR make accurate and robust 3D object detection even more difficult. In this paper, we present a novel transformer network POAT-Net for 3D point cloud object detection. Transformer is credited with the great success in Natural Language Processing (NLP) and exhibiting inspiring potentials in point cloud processing. Our method POAT-Net is inherently insensitive to element permutations within the unordered point cloud. The associations between local points contribute significantly to 3D object detection or other 3D tasks. Parallel offset-attention is leveraged to highlight and capture subtle associations between local points. To overcome the non-uniform density distribution of different objects, we exploit Normalized multi-resolution Grouping (NMRG) strategy to enhance the non-uniform density adaptive ability for POAT-Net. Quantitative experimental results on KITTI3D dataset demonstrate our method achieves the state-of-the-art performance.Jinyang WangXiao LinHongying YuIEEEarticle3D object detectionnon-uniform densityparallel offset-attentionpoint cloudtransformerElectrical engineering. Electronics. Nuclear engineeringTK1-9971ENIEEE Access, Vol 9, Pp 151110-151117 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	3D object detection non-uniform density parallel offset-attention point cloud transformer Electrical engineering. Electronics. Nuclear engineering TK1-9971
spellingShingle	3D object detection non-uniform density parallel offset-attention point cloud transformer Electrical engineering. Electronics. Nuclear engineering TK1-9971 Jinyang Wang Xiao Lin Hongying Yu POAT-Net: Parallel Offset-Attention Assisted Transformer for 3D Object Detection for Autonomous Driving
description	3D object detection is playing a key role in the perception process of autonomous driving and industrial robots automation. Inherent characteristics of point cloud raise an enormous challenge to both spatial representation and association analysis. Unordered point cloud spatial data structure and density variations caused by gradually varying distances to LiDAR make accurate and robust 3D object detection even more difficult. In this paper, we present a novel transformer network POAT-Net for 3D point cloud object detection. Transformer is credited with the great success in Natural Language Processing (NLP) and exhibiting inspiring potentials in point cloud processing. Our method POAT-Net is inherently insensitive to element permutations within the unordered point cloud. The associations between local points contribute significantly to 3D object detection or other 3D tasks. Parallel offset-attention is leveraged to highlight and capture subtle associations between local points. To overcome the non-uniform density distribution of different objects, we exploit Normalized multi-resolution Grouping (NMRG) strategy to enhance the non-uniform density adaptive ability for POAT-Net. Quantitative experimental results on KITTI3D dataset demonstrate our method achieves the state-of-the-art performance.
format	article
author	Jinyang Wang Xiao Lin Hongying Yu
author_facet	Jinyang Wang Xiao Lin Hongying Yu
author_sort	Jinyang Wang
title	POAT-Net: Parallel Offset-Attention Assisted Transformer for 3D Object Detection for Autonomous Driving
title_short	POAT-Net: Parallel Offset-Attention Assisted Transformer for 3D Object Detection for Autonomous Driving
title_full	POAT-Net: Parallel Offset-Attention Assisted Transformer for 3D Object Detection for Autonomous Driving
title_fullStr	POAT-Net: Parallel Offset-Attention Assisted Transformer for 3D Object Detection for Autonomous Driving
title_full_unstemmed	POAT-Net: Parallel Offset-Attention Assisted Transformer for 3D Object Detection for Autonomous Driving
title_sort	poat-net: parallel offset-attention assisted transformer for 3d object detection for autonomous driving
publisher	IEEE
publishDate	2021
url	https://doaj.org/article/37203ad23a904e6aa2175a9f34c8069d
work_keys_str_mv	AT jinyangwang poatnetparalleloffsetattentionassistedtransformerfor3dobjectdetectionforautonomousdriving AT xiaolin poatnetparalleloffsetattentionassistedtransformerfor3dobjectdetectionforautonomousdriving AT hongyingyu poatnetparalleloffsetattentionassistedtransformerfor3dobjectdetectionforautonomousdriving
_version_	1718426040748998656

POAT-Net: Parallel Offset-Attention Assisted Transformer for 3D Object Detection for Autonomous Driving

Ejemplares similares