Effective Rotation-invariant Point CNN with Spherical Harmonics kernels

Abstract : We present a novel rotation invariant architecture operating directly on point cloud data. We demonstrate how rotation invariance can be injected into a recently proposed point-based PCNN architecture, at all layers of the network, achieving invariance to both global shape transformations, and to local rotations on the level of patches or parts, useful when dealing with non-rigid objects. We achieve this by employing a spherical harmonics based kernel at different layers of the network, which is guaranteed to be invariant to rigid motions. We also introduce a more efficient pooling operation for PCNN using space-partitioning data-structures. This results in a flexible, simple and efficient architecture that achieves accurate results on challenging shape analysis tasks including classification and segmen-tation, without requiring data-augmentation, typically employed by non-invariant approaches.
Document type :
Preprints, Working Papers, ...
Complete list of metadatas

Cited literature [31 references]  Display  Hide  Download

https://hal.inria.fr/hal-02167454
Contributor : Yann Ponty <>
Submitted on : Thursday, June 27, 2019 - 7:23:15 PM
Last modification on : Monday, July 8, 2019 - 2:59:19 PM

File

main.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-02167454, version 1

Citation

Adrien Poulenard, Marie-Julie Rakotosaona, Yann Ponty, Maks Ovsjanikov. Effective Rotation-invariant Point CNN with Spherical Harmonics kernels. 2019. ⟨hal-02167454⟩

Share

Metrics

Record views

53

Files downloads

335