Details
| Originalsprache | Englisch |
|---|---|
| Titel des Sammelwerks | 2022 IEEE 25th International Conference on Intelligent Transportation Systems, ITSC 2022 |
| Seiten | 111-117 |
| Seitenumfang | 7 |
| ISBN (elektronisch) | 978-1-6654-6880-0 |
| Publikationsstatus | Veröffentlicht - 1 Nov. 2022 |
| Veranstaltung | 25th IEEE International Conference on Intelligent Transportation Systems, ITSC 2022 - Macau, China Dauer: 8 Okt. 2022 → 12 Okt. 2022 |
Abstract
This paper presents novel hybrid architectures that combine grid- and point-based processing to improve the detection performance and orientation estimation of radar-based object detection networks. Purely grid-based detection models operate on a bird's-eye-view (BEV) projection of the input point cloud. These approaches suffer from a loss of detailed information through the discrete grid resolution. This applies in particular to radar object detection, where relatively coarse grid resolutions are commonly used to account for the sparsity of radar point clouds. In contrast, point-based models are not affected by this problem as they process point clouds without discretization. However, they generally exhibit worse detection performances than grid-based methods. We show that a point-based model can extract neighborhood features, leveraging the exact relative positions of points, before grid rendering. This has significant benefits for a subsequent grid-based convolutional detection backbone. In experiments on the public nuScenes dataset our hybrid architecture achieves improvements in terms of detection performance (19.7% higher mAP for car class than next-best radar-only submission) and orientation estimates (11.5% relative orientation improvement) over networks from previous literature.
ASJC Scopus Sachgebiete
- Ingenieurwesen (insg.)
- Maschinenbau
- Ingenieurwesen (insg.)
- Fahrzeugbau
- Informatik (insg.)
- Angewandte Informatik
Zitieren
- Standard
- Harvard
- Apa
- Vancouver
- BibTex
- RIS
2022 IEEE 25th International Conference on Intelligent Transportation Systems, ITSC 2022. 2022. S. 111-117.
Publikation: Beitrag in Buch/Bericht/Sammelwerk/Konferenzband › Aufsatz in Konferenzband › Forschung › Peer-Review
}
TY - GEN
T1 - Improved Orientation Estimation and Detection with Hybrid Object Detection Networks for Automotive Radar
AU - Ulrich, Michael
AU - Braun, Sascha
AU - Köhler, Daniel
AU - Niederlöhner, Daniel
AU - Faion, Florian
AU - Gläser, Claudius
AU - Blume, Holger
N1 - Publisher Copyright: © 2022 IEEE.
PY - 2022/11/1
Y1 - 2022/11/1
N2 - This paper presents novel hybrid architectures that combine grid- and point-based processing to improve the detection performance and orientation estimation of radar-based object detection networks. Purely grid-based detection models operate on a bird's-eye-view (BEV) projection of the input point cloud. These approaches suffer from a loss of detailed information through the discrete grid resolution. This applies in particular to radar object detection, where relatively coarse grid resolutions are commonly used to account for the sparsity of radar point clouds. In contrast, point-based models are not affected by this problem as they process point clouds without discretization. However, they generally exhibit worse detection performances than grid-based methods. We show that a point-based model can extract neighborhood features, leveraging the exact relative positions of points, before grid rendering. This has significant benefits for a subsequent grid-based convolutional detection backbone. In experiments on the public nuScenes dataset our hybrid architecture achieves improvements in terms of detection performance (19.7% higher mAP for car class than next-best radar-only submission) and orientation estimates (11.5% relative orientation improvement) over networks from previous literature.
AB - This paper presents novel hybrid architectures that combine grid- and point-based processing to improve the detection performance and orientation estimation of radar-based object detection networks. Purely grid-based detection models operate on a bird's-eye-view (BEV) projection of the input point cloud. These approaches suffer from a loss of detailed information through the discrete grid resolution. This applies in particular to radar object detection, where relatively coarse grid resolutions are commonly used to account for the sparsity of radar point clouds. In contrast, point-based models are not affected by this problem as they process point clouds without discretization. However, they generally exhibit worse detection performances than grid-based methods. We show that a point-based model can extract neighborhood features, leveraging the exact relative positions of points, before grid rendering. This has significant benefits for a subsequent grid-based convolutional detection backbone. In experiments on the public nuScenes dataset our hybrid architecture achieves improvements in terms of detection performance (19.7% higher mAP for car class than next-best radar-only submission) and orientation estimates (11.5% relative orientation improvement) over networks from previous literature.
KW - cs.CV
KW - cs.AI
KW - cs.LG
KW - cs.RO
UR - http://www.scopus.com/inward/record.url?scp=85141826783&partnerID=8YFLogxK
U2 - 10.1109/ITSC55140.2022.9922457
DO - 10.1109/ITSC55140.2022.9922457
M3 - Conference contribution
SN - 978-1-6654-6881-7
SP - 111
EP - 117
BT - 2022 IEEE 25th International Conference on Intelligent Transportation Systems, ITSC 2022
T2 - 25th IEEE International Conference on Intelligent Transportation Systems, ITSC 2022
Y2 - 8 October 2022 through 12 October 2022
ER -