3D-MiniNet: Learning a 2D Representation From Point Clouds for Fast and Efficient 3D LIDAR Semantic Segmentation

Alonso, Íñigo; Riazuelo, Luis; Montesano, Luis; Murillo, Ana C.

doi:10.1109/lra.2020.3007440

Cited by 127 publications

(67 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…AMVNet [45] considered the class-prediction uncertainty between different view's networks, and these uncertain points are passed into an extra network to obtain more robust predictions. 3D-MiniNet [46] learns a 2D representation from the raw points through a projection which extracts local and global information from the 3D data, and then feeds it to a 2D fully convolutional neural network to generate semantic predictions.…”

Section: ) Deep Learning Based On Regular Representationsmentioning

confidence: 99%

A Comparison of Deep Learning Methods for Airborne Lidar Point Clouds Classification

Kähler

Pfeifer

2021

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

View full text Add to dashboard Cite

The success achieved by deep learning techniques in image labelling has triggered a growing interest in applying deep learning for 3D point cloud classification. To provide better insights into different deep learning architectures and their applications to ALS point cloud classification, this paper presents a comprehensive comparison among three state-of-theart deep learning networks: PointNet++, SparseCNN and KPConv, on two different ALS datasets. The performances of these three deep learning networks are compared w.r.t. classification accuracy, computation time, generalization ability as well as the sensitivity to the choices of hyper-parameters. Overall, we observed that PointNet++, SparseCNN and KPConv all outperform Random Forest on the classification results. Moreover, SparseCNN leads to a slightly better classification result compared to PointNet++ and KPConv, while requiring less computation time and memory. At the same time, it shows a better ability to generalize and is less impacted by the different choices of hyper-parameters.

show abstract

Section: ) Deep Learning Based On Regular Representationsmentioning

confidence: 99%

A Comparison of Deep Learning Methods for Airborne Lidar Point Clouds Classification

Kähler

Pfeifer

2021

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

View full text Add to dashboard Cite

show abstract

“…Due to the irregularities of the point cloud, many 3D detection algorithms convert the point cloud into 2D images to extract features in a 2D convolution manner and detect the method [18][19][20]. Zeng et al [21] proposed a Real-Time 3D (RT3D) algorithm for converting 3D point clouds into 2D grids.…”

Section: Lidarmentioning

confidence: 99%

[Retracted] Sensor‐Based Environmental Perception Technology for Intelligent Vehicles

et al. 2021

View full text Add to dashboard Cite

Environmental perception technology is the basis and premise of intelligent vehicle decision control of intelligent vehicles, a crucial link of intelligent vehicles to realize intelligence, and also the basic guarantee of its safety and intelligence. The accuracy and robustness of the perception algorithm will directly affect or even determine the realization of the upper function of intelligent vehicles. The wrong environmental perception will affect the control of the vehicle, thus causing safety risks. This paper discusses the intelligent vehicle perception technology and introduces the development status and control strategies of several important sensors such as machine vision, laser radar, and millimeter-wave radar. Target detection, target recognition, and multisensor fusion are analyzed in the optimized part of sensor results. The functions of the intelligent vehicle assistance system which has been applied to the ground at present are described, and the lane detection, adaptive cruise control (ACC), and autonomous emergency braking (AEB) are analyzed. Finally, the paper looks forward to the research direction of sense-based intelligent vehicle perception technology, which will play an important role in guiding the development of intelligent vehicles and accelerate the landing process of intelligent vehicles.

show abstract

“…The 3D-Mininet proposed by Alonso obtained local and global context information from 3D data through multiview projection, and then input these data to a 2D full convolutional neural network (FCNN) to predict semantic labels. Finally, the predicted 2D semantic labels were reprojected into 3D space to obtain high-quality results in a faster and more efficient way [30].…”

Section: Two-dimensional Multiview Methodsmentioning

confidence: 99%

Integrating Normal Vector Features into an Atrous Convolution Residual Network for LiDAR Point Cloud Classification

Zhang

Jiang

et al. 2021

Remote Sensing

View full text Add to dashboard Cite

LiDAR point clouds are rich in spatial information and can effectively express the size, shape, position, and direction of objects; thus, they have the advantage of high spatial utilization. The point cloud focuses on describing the shape of the external surface of the object itself and will not store useless redundant information to describe the occupation. Therefore, point clouds have become the research focus of 3D data models and are widely used in large-scale scene reconstruction, virtual reality, digital elevation model production, and other fields. Since point clouds have various characteristics, such as disorder, density inconsistency, unstructuredness, and incomplete information, point cloud classification is still complex and challenging. To realize the semantic classification of LiDAR point clouds in complex scenarios, this paper proposes the integration of normal vector features into an atrous convolution residual network. Based on the RandLA-Net network structure, the proposed network integrates the atrous convolution into the residual module to extract global and local features of the point clouds. The atrous convolution can learn more valuable point cloud feature information by expanding the receptive field. Then, the point cloud normal vector is embedded in the local feature aggregation module of the RandLA-Net network to extract local semantic aggregation features. The improved local feature aggregation module can merge the deep features of the point cloud and mine the fine-grained information of the point cloud to improve the model's segmentation ability in complex scenes. Finally, to resolve the imbalance of the distribution of the various categories of point clouds, the original loss function is optimized by adopting a reweighted method to prevent overfitting so that the network can focus on small target categories in the training process to effectively improve the classification performance. Through the experimental analysis of a Vaihingen (Germany) urban 3D semantic dataset from the ISPRS website, it is verified that the proposed algorithm has a strong generalization ability. The overall accuracy (OA) of the proposed algorithm on the Vaihingen urban 3D semantic dataset reached 97.9%, and the average reached 96.1%. Experiments show that the proposed algorithm fully exploits the semantic features of point clouds and effectively improves the accuracy of point cloud classification.

show abstract

3D-MiniNet: Learning a 2D Representation From Point Clouds for Fast and Efficient 3D LIDAR Semantic Segmentation

Cited by 127 publications

References 29 publications

A Comparison of Deep Learning Methods for Airborne Lidar Point Clouds Classification

A Comparison of Deep Learning Methods for Airborne Lidar Point Clouds Classification

[Retracted] Sensor‐Based Environmental Perception Technology for Intelligent Vehicles

Integrating Normal Vector Features into an Atrous Convolution Residual Network for LiDAR Point Cloud Classification

Contact Info

Product

Resources

About