Aggregated Time Series Features in a Voxel-Based Network Architecture

Using point cloud sequences is a popular way to harness the additional information represented in the time domain in order to enhance the performance of 3D object detector neural networks. However, it is not trivial to decide which abstraction level should the additional information presented to the...

Full description

Saved in:
Bibliographic Details
Main Authors: Zsolt Vincze, Andras Rovid
Format: Article
Language:English
Published: IEEE 2025-01-01
Series:IEEE Access
Subjects:
Online Access:https://ieeexplore.ieee.org/document/10855412/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1832575562494771200
author Zsolt Vincze
Andras Rovid
author_facet Zsolt Vincze
Andras Rovid
author_sort Zsolt Vincze
collection DOAJ
description Using point cloud sequences is a popular way to harness the additional information represented in the time domain in order to enhance the performance of 3D object detector neural networks. However, it is not trivial to decide which abstraction level should the additional information presented to the network, or what is the point in the architecture, where aggregating the additional information is most beneficial. In this article, the authors propose various voxel-based networks and analyze their performance in relation to the abstraction level of the time series data. During the evaluation, the authors examine the object detection performance of a popular voxel-based neural network with its original architecture and several variants where the time domain related features were propagated through the network and aggregated at different stages of processing. Based on the evaluation results, a conclusion is drawn regarding the abstraction level at which the time-series aggregation step is performed in order to improve the performance of the baseline voxel-based detector.
format Article
id doaj-art-041371634e8444ff9ff518b992fdd723
institution Kabale University
issn 2169-3536
language English
publishDate 2025-01-01
publisher IEEE
record_format Article
series IEEE Access
spelling doaj-art-041371634e8444ff9ff518b992fdd7232025-01-31T23:04:44ZengIEEEIEEE Access2169-35362025-01-0113193291933910.1109/ACCESS.2025.353515110855412Aggregated Time Series Features in a Voxel-Based Network ArchitectureZsolt Vincze0https://orcid.org/0000-0002-5813-3530Andras Rovid1https://orcid.org/0000-0002-9044-1760Department of Automotive Technologies, Faculty of Transportation Engineering and Vehicle Engineering, Budapest University of Technology and Economics, Budapest, HungaryDepartment of Automotive Technologies, Faculty of Transportation Engineering and Vehicle Engineering, Budapest University of Technology and Economics, Budapest, HungaryUsing point cloud sequences is a popular way to harness the additional information represented in the time domain in order to enhance the performance of 3D object detector neural networks. However, it is not trivial to decide which abstraction level should the additional information presented to the network, or what is the point in the architecture, where aggregating the additional information is most beneficial. In this article, the authors propose various voxel-based networks and analyze their performance in relation to the abstraction level of the time series data. During the evaluation, the authors examine the object detection performance of a popular voxel-based neural network with its original architecture and several variants where the time domain related features were propagated through the network and aggregated at different stages of processing. Based on the evaluation results, a conclusion is drawn regarding the abstraction level at which the time-series aggregation step is performed in order to improve the performance of the baseline voxel-based detector.https://ieeexplore.ieee.org/document/10855412/Neural networksLiDAR point cloudpoint cloud sequencetime seriesobject detection
spellingShingle Zsolt Vincze
Andras Rovid
Aggregated Time Series Features in a Voxel-Based Network Architecture
IEEE Access
Neural networks
LiDAR point cloud
point cloud sequence
time series
object detection
title Aggregated Time Series Features in a Voxel-Based Network Architecture
title_full Aggregated Time Series Features in a Voxel-Based Network Architecture
title_fullStr Aggregated Time Series Features in a Voxel-Based Network Architecture
title_full_unstemmed Aggregated Time Series Features in a Voxel-Based Network Architecture
title_short Aggregated Time Series Features in a Voxel-Based Network Architecture
title_sort aggregated time series features in a voxel based network architecture
topic Neural networks
LiDAR point cloud
point cloud sequence
time series
object detection
url https://ieeexplore.ieee.org/document/10855412/
work_keys_str_mv AT zsoltvincze aggregatedtimeseriesfeaturesinavoxelbasednetworkarchitecture
AT andrasrovid aggregatedtimeseriesfeaturesinavoxelbasednetworkarchitecture