LU-Net: An Efficient Network for 3D LiDAR Point Cloud Semantic Segmentation Based on End-to-End-Learned 3D Features and U-Net

Pierre Biasutti, Vincent Lepetit, Jean-Francois Aujol, Mathieu Brédif, Aurélie Bugeau

Research output: Chapter in Book/Report/Conference proceedingConference contributionResearchpeer-review

Abstract

We propose LU-Net-for LiDAR U-Net, a new method for the semantic segmentation of a 3D LiDAR point cloud. Instead of applying some global 3D segmentation method such as PointNet, we propose an end-to-end architecture for LiDAR point cloud semantic segmentation that efficiently solves the problem as an image processing problem. We first extract high-level 3D features for each point given its 3D neighbors. Then, these features are projected into a 2D multichannel range-image by considering the topology of the sensor. Thanks to these learned features and this projection, we can finally perform the segmentation using a simple U-Net segmentation network, which performs very well while being very efficient. In this way, we can exploit both the 3D nature of the data and the specificity of the LiDAR sensor. This approach outperforms the state-of-the-art by a large margin on the KITTI dataset, as our experiments show. Moreover, this approach operates at 24fps on a single GPU. This is above the acquisition rate of common LiDAR sensors which makes it suitable for real-time applications.
Original languageEnglish
Title of host publicationProceedings of the IEEE International Conference on Computer Vision Workshops
Publication statusPublished - 2019

Fingerprint

Semantics
Sensors
Image processing
Topology
Experiments
Graphics processing unit

Cite this

Biasutti, P., Lepetit, V., Aujol, J-F., Brédif, M., & Bugeau, A. (2019). LU-Net: An Efficient Network for 3D LiDAR Point Cloud Semantic Segmentation Based on End-to-End-Learned 3D Features and U-Net. In Proceedings of the IEEE International Conference on Computer Vision Workshops

LU-Net: An Efficient Network for 3D LiDAR Point Cloud Semantic Segmentation Based on End-to-End-Learned 3D Features and U-Net. / Biasutti, Pierre; Lepetit, Vincent; Aujol, Jean-Francois; Brédif, Mathieu; Bugeau, Aurélie.

Proceedings of the IEEE International Conference on Computer Vision Workshops. 2019.

Research output: Chapter in Book/Report/Conference proceedingConference contributionResearchpeer-review

Biasutti, P, Lepetit, V, Aujol, J-F, Brédif, M & Bugeau, A 2019, LU-Net: An Efficient Network for 3D LiDAR Point Cloud Semantic Segmentation Based on End-to-End-Learned 3D Features and U-Net. in Proceedings of the IEEE International Conference on Computer Vision Workshops.
Biasutti P, Lepetit V, Aujol J-F, Brédif M, Bugeau A. LU-Net: An Efficient Network for 3D LiDAR Point Cloud Semantic Segmentation Based on End-to-End-Learned 3D Features and U-Net. In Proceedings of the IEEE International Conference on Computer Vision Workshops. 2019
Biasutti, Pierre ; Lepetit, Vincent ; Aujol, Jean-Francois ; Brédif, Mathieu ; Bugeau, Aurélie. / LU-Net: An Efficient Network for 3D LiDAR Point Cloud Semantic Segmentation Based on End-to-End-Learned 3D Features and U-Net. Proceedings of the IEEE International Conference on Computer Vision Workshops. 2019.
@inproceedings{7e05f03e3ace4e0ab5eefc1b9fd0aa96,
title = "LU-Net: An Efficient Network for 3D LiDAR Point Cloud Semantic Segmentation Based on End-to-End-Learned 3D Features and U-Net",
abstract = "We propose LU-Net-for LiDAR U-Net, a new method for the semantic segmentation of a 3D LiDAR point cloud. Instead of applying some global 3D segmentation method such as PointNet, we propose an end-to-end architecture for LiDAR point cloud semantic segmentation that efficiently solves the problem as an image processing problem. We first extract high-level 3D features for each point given its 3D neighbors. Then, these features are projected into a 2D multichannel range-image by considering the topology of the sensor. Thanks to these learned features and this projection, we can finally perform the segmentation using a simple U-Net segmentation network, which performs very well while being very efficient. In this way, we can exploit both the 3D nature of the data and the specificity of the LiDAR sensor. This approach outperforms the state-of-the-art by a large margin on the KITTI dataset, as our experiments show. Moreover, this approach operates at 24fps on a single GPU. This is above the acquisition rate of common LiDAR sensors which makes it suitable for real-time applications.",
author = "Pierre Biasutti and Vincent Lepetit and Jean-Francois Aujol and Mathieu Br{\'e}dif and Aur{\'e}lie Bugeau",
year = "2019",
language = "English",
booktitle = "Proceedings of the IEEE International Conference on Computer Vision Workshops",

}

TY - GEN

T1 - LU-Net: An Efficient Network for 3D LiDAR Point Cloud Semantic Segmentation Based on End-to-End-Learned 3D Features and U-Net

AU - Biasutti, Pierre

AU - Lepetit, Vincent

AU - Aujol, Jean-Francois

AU - Brédif, Mathieu

AU - Bugeau, Aurélie

PY - 2019

Y1 - 2019

N2 - We propose LU-Net-for LiDAR U-Net, a new method for the semantic segmentation of a 3D LiDAR point cloud. Instead of applying some global 3D segmentation method such as PointNet, we propose an end-to-end architecture for LiDAR point cloud semantic segmentation that efficiently solves the problem as an image processing problem. We first extract high-level 3D features for each point given its 3D neighbors. Then, these features are projected into a 2D multichannel range-image by considering the topology of the sensor. Thanks to these learned features and this projection, we can finally perform the segmentation using a simple U-Net segmentation network, which performs very well while being very efficient. In this way, we can exploit both the 3D nature of the data and the specificity of the LiDAR sensor. This approach outperforms the state-of-the-art by a large margin on the KITTI dataset, as our experiments show. Moreover, this approach operates at 24fps on a single GPU. This is above the acquisition rate of common LiDAR sensors which makes it suitable for real-time applications.

AB - We propose LU-Net-for LiDAR U-Net, a new method for the semantic segmentation of a 3D LiDAR point cloud. Instead of applying some global 3D segmentation method such as PointNet, we propose an end-to-end architecture for LiDAR point cloud semantic segmentation that efficiently solves the problem as an image processing problem. We first extract high-level 3D features for each point given its 3D neighbors. Then, these features are projected into a 2D multichannel range-image by considering the topology of the sensor. Thanks to these learned features and this projection, we can finally perform the segmentation using a simple U-Net segmentation network, which performs very well while being very efficient. In this way, we can exploit both the 3D nature of the data and the specificity of the LiDAR sensor. This approach outperforms the state-of-the-art by a large margin on the KITTI dataset, as our experiments show. Moreover, this approach operates at 24fps on a single GPU. This is above the acquisition rate of common LiDAR sensors which makes it suitable for real-time applications.

M3 - Conference contribution

BT - Proceedings of the IEEE International Conference on Computer Vision Workshops

ER -