Papers › FG-Net: Fast Large-Scale LiDAR Point Clouds Understanding Network Leveraging...

FG-Net: Fast Large-Scale LiDAR Point Clouds Understanding Network Leveraging Correlated Feature Mining and Geometric-Aware Modelling

17 Dec 2020arXiv:2012.09439archive 2025-07-28

Kangcheng Liu, Zhi Gao, Feng Lin, Ben M. Chen

This work presents FG-Net, a general deep learning framework for large-scale point clouds understanding without voxelizations, which achieves accurate and real-time performance with a single NVIDIA GTX 1080 GPU. First, a novel noise and outlier filtering method is designed to facilitate subsequent high-level tasks. For effective understanding purpose, we propose a deep convolutional neural network leveraging correlated feature mining and deformable convolution based geometric-aware modelling, in which the local feature relationships and geometric patterns can be fully exploited. For the efficiency issue, we put forward an inverse density sampling operation and a feature pyramid based residual learning strategy to save the computational cost and memory consumption respectively. Extensive experiments on real-world challenging datasets demonstrated that our approaches outperform state-of-the-art approaches in terms of accuracy and efficiency. Moreover, weakly supervised transfer learning is also conducted to demonstrate the generalization capacity of our method.

PaperPDFCode

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

Code

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

3D Part Segmentation3D Point Cloud Classification3D Semantic SegmentationLIDAR Semantic SegmentationSemantic SegmentationTransfer LearningWeakly supervised segmentation

1 archive task tag without a task page not shown.

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
3D Part Segmentation ShapeNet-Part Feature Geometric Net (FG-Net) Class Average IoU 87.7 #17 of 67 Archive leaderboard report
3D Part Segmentation ShapeNet-Part Feature Geometric Net (FG-Net) Instance Average IoU 86.6 #17 of 67 Archive leaderboard report
3D Point Cloud Classification ModelNet40 Feature Geometric Net (FG-Net) Mean Accuracy 91.1 #40 of 111 Archive leaderboard report
3D Point Cloud Classification ModelNet40 Feature Geometric Net (FG-Net) Overall Accuracy 93.8 #40 of 111 Archive leaderboard report
3D Semantic Segmentation PartNet FG-Net mIOU 58.2 #3 of 6 Archive leaderboard report
3D Semantic Segmentation SemanticKITTI FG-Net test mIoU 53.8% #31 of 45 Archive leaderboard report
LIDAR Semantic Segmentation Paris-Lille-3D Feature Geometric Net (FG Net) mIOU 0.819 #2 of 9 Archive leaderboard report
Semantic Segmentation S3DIS Feature Geometric Net (FG-Net) Mean IoU 70.8 #21 of 54 Archive leaderboard report
Semantic Segmentation S3DIS Feature Geometric Net (FG-Net) Number of params N/A #21 of 54 Archive leaderboard report
Semantic Segmentation S3DIS Feature Geometric Net (FG-Net) mAcc 82.9 #21 of 54 Archive leaderboard report
Semantic Segmentation S3DIS Feature Geometric Net (FG-Net) oAcc 88.2 #21 of 54 Archive leaderboard report
Semantic Segmentation ScanNet FG-Net test mIoU 69.0 #34 of 45 Archive leaderboard report
Semantic Segmentation Semantic3D Feature Geometric Net mIoU 78.2% #1 of 17 Archive leaderboard report
Semantic Segmentation Semantic3D Feature Geometric Net oAcc 93.6 #1 of 17 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

ConvolutionDeformable Convolution

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections