Papers › PointCNN: Convolution On X-Transformed Points

PointCNN: Convolution On X-Transformed Points

1 Dec 2018NeurIPS 2018 12archive 2025-07-28

Yangyan Li, Rui Bu, Mingchao Sun, Wei Wu, Xinhan Di, Baoquan Chen

We present a simple and general framework for feature learning from point cloud. The key to the success of CNNs is the convolution operator that is capable of leveraging spatially-local correlation in data represented densely in grids (e.g. images). However, point cloud are irregular and unordered, thus a direct convolving of kernels against the features associated with the points will result in deserting the shape information while being variant to the orders. To address these problems, we propose to learn a X-transformation from the input points, which is used for simultaneously weighting the input features associated with the points and permuting them into latent potentially canonical order. Then element-wise product and sum operations of typical convolution operator are applied on the X-transformed features. The proposed method is a generalization of typical CNNs into learning features from point cloud, thus we call it PointCNN. Experiments show that PointCNN achieves on par or better performance than state-of-the-art methods on multiple challenging benchmark datasets and tasks.

PaperPDFCode

Code

yangyanli/PointCNN officialmentioned in papertfNOASSERTION report

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 SegmentationFew-Shot 3D Point Cloud ClassificationSemantic Segmentation

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
3D Point Cloud Classification IntrA PointCNN F1 score (5-fold) 0.875 #6 of 12 Archive leaderboard report
3D Point Cloud Classification ModelNet40 PointCNN Overall Accuracy 92.2 #91 of 111 Archive leaderboard report
3D Semantic Segmentation DALES PointCNN Model size N/A #8 of 9 Archive leaderboard report
3D Semantic Segmentation DALES PointCNN Overall Accuracy 97.2 #8 of 9 Archive leaderboard report
3D Semantic Segmentation DALES PointCNN mIoU 58.4 #8 of 9 Archive leaderboard report
Few-Shot 3D Point Cloud Classification ModelNet40 10-way (10-shot) PointCNN Overall Accuracy 46.60 #28 of 31 Archive leaderboard report
Few-Shot 3D Point Cloud Classification ModelNet40 10-way (10-shot) PointCNN Standard Deviation 4.8 #28 of 31 Archive leaderboard report
Few-Shot 3D Point Cloud Classification ModelNet40 10-way (20-shot) PointCNN Overall Accuracy 49.95 #28 of 31 Archive leaderboard report
Few-Shot 3D Point Cloud Classification ModelNet40 10-way (20-shot) PointCNN Standard Deviation 7.2 #28 of 31 Archive leaderboard report
Few-Shot 3D Point Cloud Classification ModelNet40 5-way (20-shot) PointCNN Overall Accuracy 68.64 #26 of 30 Archive leaderboard report
Few-Shot 3D Point Cloud Classification ModelNet40 5-way (20-shot) PointCNN Standard Deviation 7.0 #26 of 30 Archive leaderboard report
Semantic Segmentation S3DIS Area5 PointCNN Number of params N/A #61 of 61 Archive leaderboard report
Semantic Segmentation S3DIS Area5 PointCNN oAcc 85.9 #61 of 61 Archive leaderboard report
Semantic Segmentation ScanNet PointCNN test mIoU 45.8 #40 of 45 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

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