Skip to search boxSkip to navigationSkip to main content

3DeformRS: Certifying Spatial Deformations on Point Clouds

  • S. Gabriel Perez
    ,
  • ,
  • Motasem Alfarra
    ,
  • Silvio Giancola
    ,
  • Bernard Ghanem
  • Universidad Nacional de Colombia
    ,
  • King Abdullah University of Science and Technology
Research Output:
Chapter in Book/Report/Conference proceeding
Conference contribution
Research
Peer-review

Abstract

3D computer vision models are commonly used in security-critical applications such as autonomous driving and surgical robotics. Emerging concerns over the robustness of these models against real-world deformations must be addressed practically and reliably. In this work, we propose 3DeformRS, a method to certify the robustness of point cloud Deep Neural Networks (DNNs) against real-world deformations. We developed 3DeformRS by building upon recent work that generalized Randomized Smoothing (RS) from pixel-intensity perturbations to vector-field deformations. In particular, we specialized RS to certify DNNs against parameterized deformations (e.g. rotation, twisting), while enjoying practical computational costs. We leverage the virtues of 3DeformRS to conduct a comprehensive empirical study on the certified robustness of four representative point cloud DNNs on two datasets and against seven different deformations. Compared to previous approaches for certifying point cloud DNNs, 3DeformRS is fast, scales well with point cloud size, and provides comparable-to-better certificates. For instance, when certifying a plain PointNet against a 3° z-rotation on 1024-point clouds, 3DeformRS grants a certificate 3× larger and 20× faster than previous work 11Code:https://github.com/gaperezsa/3DeformRS.

Research Output Information

Output type

Research Output:
Chapter in Book/Report/Conference proceeding
Conference contribution
Research
Peer-review

Original language

English

Pages from-to (Number of pages)

Pages 15148-15158 (11 pages)

Publication milestones

  • Published - 2022

Publication status

Published - 2022

Publisher

IEEE Computer Society, United States

Publication series

  • Publication series name: Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition
    ISSN (Print): 1063-6919
    Volume: 2022-June

ISBN (Electronic)

9781665469463

Publication IDs

  • Scopus: 85140199096

Host publication title

Proceedings - 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition, CVPR 2022

Funding Details

Acknowledgements. This publication is based upon work supported by the King Abdullah University of Science and Technology (KAUST) Office of Sponsored Research (OSR) under Award No. OSR-CRG2019-4033. We would also like to thank Jesús Zarzar for the help and discussions.
FunderFunding number
KAUST
OSR-CRG2019-4033

Publication metrics

Metrics

Scopus
citations
SciVal
FWCI
0.54
SciVal
Author count
5
SciVal
Paper percentile
53
SciVal
citations
10

PlumX, opens in new tab

Citation count
10
Captures
37