Home > Research > Publications & Outputs > Deep Supervised Descent Method With Multiple Se...

Links

Text available via DOI:

View graph of relations

Deep Supervised Descent Method With Multiple Seeds Generation for 3-D Tracking in Point Cloud

Research output: Contribution to Journal/MagazineJournal articlepeer-review

Published

Standard

Deep Supervised Descent Method With Multiple Seeds Generation for 3-D Tracking in Point Cloud. / Tian, Shengjing; Liu, Bin; Tan, Hongchen et al.
In: IEEE Transactions on Industrial Informatics, Vol. 18, No. 8, 31.08.2022, p. 5077-5086.

Research output: Contribution to Journal/MagazineJournal articlepeer-review

Harvard

Tian, S, Liu, B, Tan, H, Liu, J, Liu, M & Liu, X 2022, 'Deep Supervised Descent Method With Multiple Seeds Generation for 3-D Tracking in Point Cloud', IEEE Transactions on Industrial Informatics, vol. 18, no. 8, pp. 5077-5086. https://doi.org/10.1109/TII.2021.3126391

APA

Tian, S., Liu, B., Tan, H., Liu, J., Liu, M., & Liu, X. (2022). Deep Supervised Descent Method With Multiple Seeds Generation for 3-D Tracking in Point Cloud. IEEE Transactions on Industrial Informatics, 18(8), 5077-5086. https://doi.org/10.1109/TII.2021.3126391

Vancouver

Tian S, Liu B, Tan H, Liu J, Liu M, Liu X. Deep Supervised Descent Method With Multiple Seeds Generation for 3-D Tracking in Point Cloud. IEEE Transactions on Industrial Informatics. 2022 Aug 31;18(8):5077-5086. Epub 2021 Nov 9. doi: 10.1109/TII.2021.3126391

Author

Tian, Shengjing ; Liu, Bin ; Tan, Hongchen et al. / Deep Supervised Descent Method With Multiple Seeds Generation for 3-D Tracking in Point Cloud. In: IEEE Transactions on Industrial Informatics. 2022 ; Vol. 18, No. 8. pp. 5077-5086.

Bibtex

@article{8b78f60120364501a43a1be803643909,
title = "Deep Supervised Descent Method With Multiple Seeds Generation for 3-D Tracking in Point Cloud",
abstract = "Three-dimensional (3-D) tracking in point cloud is a core competence of autonomous robots to perceive and forecast the environment. How to initialize bounding box seeds and optimize their position and orientation are very crucial for 3-D object tracking in point clouds. Nevertheless, existing methods mainly resort to developing a powerful classifier based on the initial bounding box seeds. In this article, we propose an end-to-end deep supervised descent method (SDM), which seamlessly integrates multiple seeds generation for the initialization of seeds and sequential updates for the estimation of accurate result. Specifically, we start with transforming the SDM iterative process into a trainable recurrent module. It explicitly learns a series of descent directions in the parameter space, to gradually optimize the initial seeds. Moreover, to alleviate drifting of this process, we initialize multiple seeds based on aggregated point sets generated by the deep Hough voting. Besides, a discrimination module is introduced to determine the bounding box with the highest score as the final result. Importantly, a specific multitask loss is proposed to train our model in an end-to-end way. Experiments on KITTI, PandaSet, and Waymo datasets show that our method could achieve significant improvements (up to 11.2% in success ratio) as compared to state-of-the-art trackers.",
author = "Shengjing Tian and Bin Liu and Hongchen Tan and Jun Liu and Meng Liu and Xiuping Liu",
year = "2022",
month = aug,
day = "31",
doi = "10.1109/TII.2021.3126391",
language = "English",
volume = "18",
pages = "5077--5086",
journal = "IEEE Transactions on Industrial Informatics",
issn = "1551-3203",
publisher = "IEEE Computer Society",
number = "8",

}

RIS

TY - JOUR

T1 - Deep Supervised Descent Method With Multiple Seeds Generation for 3-D Tracking in Point Cloud

AU - Tian, Shengjing

AU - Liu, Bin

AU - Tan, Hongchen

AU - Liu, Jun

AU - Liu, Meng

AU - Liu, Xiuping

PY - 2022/8/31

Y1 - 2022/8/31

N2 - Three-dimensional (3-D) tracking in point cloud is a core competence of autonomous robots to perceive and forecast the environment. How to initialize bounding box seeds and optimize their position and orientation are very crucial for 3-D object tracking in point clouds. Nevertheless, existing methods mainly resort to developing a powerful classifier based on the initial bounding box seeds. In this article, we propose an end-to-end deep supervised descent method (SDM), which seamlessly integrates multiple seeds generation for the initialization of seeds and sequential updates for the estimation of accurate result. Specifically, we start with transforming the SDM iterative process into a trainable recurrent module. It explicitly learns a series of descent directions in the parameter space, to gradually optimize the initial seeds. Moreover, to alleviate drifting of this process, we initialize multiple seeds based on aggregated point sets generated by the deep Hough voting. Besides, a discrimination module is introduced to determine the bounding box with the highest score as the final result. Importantly, a specific multitask loss is proposed to train our model in an end-to-end way. Experiments on KITTI, PandaSet, and Waymo datasets show that our method could achieve significant improvements (up to 11.2% in success ratio) as compared to state-of-the-art trackers.

AB - Three-dimensional (3-D) tracking in point cloud is a core competence of autonomous robots to perceive and forecast the environment. How to initialize bounding box seeds and optimize their position and orientation are very crucial for 3-D object tracking in point clouds. Nevertheless, existing methods mainly resort to developing a powerful classifier based on the initial bounding box seeds. In this article, we propose an end-to-end deep supervised descent method (SDM), which seamlessly integrates multiple seeds generation for the initialization of seeds and sequential updates for the estimation of accurate result. Specifically, we start with transforming the SDM iterative process into a trainable recurrent module. It explicitly learns a series of descent directions in the parameter space, to gradually optimize the initial seeds. Moreover, to alleviate drifting of this process, we initialize multiple seeds based on aggregated point sets generated by the deep Hough voting. Besides, a discrimination module is introduced to determine the bounding box with the highest score as the final result. Importantly, a specific multitask loss is proposed to train our model in an end-to-end way. Experiments on KITTI, PandaSet, and Waymo datasets show that our method could achieve significant improvements (up to 11.2% in success ratio) as compared to state-of-the-art trackers.

U2 - 10.1109/TII.2021.3126391

DO - 10.1109/TII.2021.3126391

M3 - Journal article

VL - 18

SP - 5077

EP - 5086

JO - IEEE Transactions on Industrial Informatics

JF - IEEE Transactions on Industrial Informatics

SN - 1551-3203

IS - 8

ER -