ConvPoint: Continuous Convolutions for Point Cloud Processing

4 Apr 2019  ·  Alexandre Boulch ·

Point clouds are unstructured and unordered data, as opposed to images. Thus, most machine learning approach developed for image cannot be directly transferred to point clouds. In this paper, we propose a generalization of discrete convolutional neural networks (CNNs) in order to deal with point clouds by replacing discrete kernels by continuous ones. This formulation is simple, allows arbitrary point cloud sizes and can easily be used for designing neural networks similarly to 2D CNNs. We present experimental results with various architectures, highlighting the flexibility of the proposed approach. We obtain competitive results compared to the state-of-the-art on shape classification, part segmentation and semantic segmentation for large-scale point clouds.

PDF Abstract
Task Dataset Model Metric Name Metric Value Global Rank Result Benchmark
3D Semantic Segmentation DALES ConvPoint mIoU 67.4 # 5
Overall Accuracy 97.2 # 3
Model size 4.7M # 6
LIDAR Semantic Segmentation Paris-Lille-3D ConvPoint mIOU 0.759 # 4
LIDAR Semantic Segmentation Paris-Lille-3D ConvPoint_Keras mIOU 0.720 # 6
Semantic Segmentation S3DIS ConvPoint Mean IoU 68.2 # 30
oAcc 88.8 # 17
Number of params 4.7M # 41
Params (M) 4.1 # 12
3D Part Segmentation ShapeNet-Part ConvPoint Class Average IoU 83.4 # 20
Instance Average IoU 85.8 # 36

Methods


No methods listed for this paper. Add relevant methods here