KPConv: Flexible and Deformable Convolution for Point Clouds

Abstract : We present Kernel Point Convolution (KPConv), a new design of point convolution, i.e. that operates on point clouds without any intermediate representation. The convolution weights of KPConv are located in Euclidean space by kernel points, and applied to the input points close to them. Its capacity to use any number of kernel points gives KPConv more flexibility than fixed grid convolutions. Furthermore, these locations are continuous in space and can be learned by the network. Therefore, KPConv can be extended to deformable convolutions that learn to adapt kernel points to local geometry. Thanks to a regular subsampling strategy, KPConv is also efficient and robust to varying densities. Whether they use deformable KPConv for complex tasks, or rigid KPconv for simpler tasks, our networks outperform state-of-the-art classification and segmentation approaches on several datasets. We also offer ablation studies and visualizations to provide understanding of what has been learned by KPConv and to validate the descriptive power of deformable KPConv.
Complete list of metadatas

https://hal.archives-ouvertes.fr/hal-02310026
Contributor : Hugues Thomas <>
Submitted on : Wednesday, October 9, 2019 - 5:59:31 PM
Last modification on : Thursday, October 10, 2019 - 1:32:24 AM

Links full text

Identifiers

  • HAL Id : hal-02310026, version 1
  • ARXIV : 1904.08889

Citation

Hugues Thomas, Charles R. Qi, Jean-Emmanuel Deschaud, Beatriz Marcotegui, Francois Goulette, et al.. KPConv: Flexible and Deformable Convolution for Point Clouds. The IEEE International Conference on Computer Vision (ICCV), Oct 2019, Séoul, South Korea. ⟨hal-02310026⟩

Share

Metrics

Record views

15