Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

Deep Parametric Continuous Convolutional Neural Networks

About

Standard convolutional neural networks assume a grid structured input is available and exploit discrete convolutions as their fundamental building blocks. This limits their applicability to many real-world applications. In this paper we propose Parametric Continuous Convolution, a new learnable operator that operates over non-grid structured data. The key idea is to exploit parameterized kernel functions that span the full continuous vector space. This generalization allows us to learn over arbitrary data structures as long as their support relationship is computable. Our experiments show significant improvement over the state-of-the-art in point cloud segmentation of indoor and outdoor scenes, and lidar motion estimation of driving scenes.

Shenlong Wang, Simon Suo, Wei-Chiu Ma, Andrei Pokrovsky, Raquel Urtasun• 2021

Related benchmarks

TaskDatasetResultRank
Semantic segmentationS3DIS (Area 5)
mIOU58.3
799
3D Point Cloud ClassificationModelNet40 (test)
OA88.9
297
Part SegmentationShapeNetPart
mIoU (Instance)85.1
198
Semantic segmentationS3DIS
mIoU58.3
88
Semantic segmentationS3DIS (test)
mIoU58.3
47
3D Semantic SegmentationS3DIS (Area 5 test (Fold #1))
mIoU58.27
19
Semantic segmentationS3DIS (Stanford Indoor Dataset)
mIoU58.27
10
Semantic segmentationDriving Scenes Dataset
Pixel Acc95.45
4
Lidar FlowDriving Scenes Dataset (test)
EPE (cm)7.81
2
Showing 9 of 9 rows

Other info

Follow for update