Skip to main content
U.S. Department of Energy
Office of Scientific and Technical Information

Transferable Adversarial Attack on 3D Object Tracking in Point Cloud

Conference ·

3D point cloud object tracking has recently witnessed considerable progress relying on deep learning. Such progress, however, mainly focuses on improving tracking accuracy. The risk, especially considering that deep neural network is vulnerable to adversarial perturbations, of a tracker being attacked is often neglected and rarely explored. In order to attract attentions to this potential risk and facilitate the study of robustness in point cloud tracking, we introduce a novel transferable attack network (TAN) to deceive 3D point cloud tracking. Specifically, TAN consists of a 3D adversarial generator, which is trained with a carefully designed multi-fold drift (MFD) loss. The MFD loss considers three common grounds, including classification, intermediate feature and angle drifts, across different 3D point cloud tracking frameworks for perturbation generation, leading to high transferability of TAN for attack. In our extensive experiments, we demonstrate the proposed TAN is able to not only drastically degrade the victim 3D point cloud tracker, \ie, P2B, but also effectively deceive other unseen state-of-the-art approaches such as BAT and M^2Track, posing a new threat to 3D point cloud tracking.

Research Organization:
Brookhaven National Laboratory (BNL), Upton, NY (United States)
Sponsoring Organization:
USDOE Office of Science (SC), Advanced Scientific Computing Research (SC-21)
DOE Contract Number:
SC0012704
OSTI ID:
1993156
Report Number(s):
BNL-224637-2023-CPPJ
Resource Relation:
Conference: International Conference on Multimedia Modeling, 2023, Bergen, Norway, 1/9/2023 - 1/12/2023
Country of Publication:
United States
Language:
English

References (24)

One-Shot Adversarial Attacks on Visual Tracking With Dual Attention conference June 2020
Siamese Cascaded Region Proposal Networks for Real-Time Visual Tracking conference June 2019
Ad2Attack: Adaptive Adversarial Attack on Real-Time UAV Tracking conference May 2022
Are we ready for autonomous driving? The KITTI vision benchmark suite conference June 2012
Leveraging Shape Completion for 3D Siamese Tracking conference June 2019
Shape-invariant 3D Adversarial Point Clouds conference June 2022
Minimal Adversarial Examples for Deep Learning on 3D Point Clouds conference October 2021
SiamRPN++: Evolution of Siamese Visual Tracking With Very Deep Networks conference June 2019
High Performance Visual Tracking with Siamese Region Proposal Network conference June 2018
Extending Adversarial Attacks and Defenses to Deep 3D Point Cloud Classifiers conference September 2019
Efficient Joint Gradient Based Attack Against SOR Defense for 3D Point Cloud Classification conference October 2020
P2B: Point-to-Box Network for 3D Object Tracking in Point Clouds conference June 2020
U-Net: Convolutional Networks for Biomedical Image Segmentation
  • Ronneberger, Olaf; Fischer, Philipp; Brox, Thomas
  • Medical Image Computing and Computer-Assisted Intervention – MICCAI 2015: 18th International Conference, Munich, Germany, October 5-9, 2015, Proceedings, Part III https://doi.org/10.1007/978-3-319-24574-4_28
book November 2015
PTT: Point-Track-Transformer Module for 3D Single Object Tracking in Point Clouds conference September 2021
MLVSNet: Multi-level Voting Siamese Network for 3D Visual Tracking conference October 2021
Physical Adversarial Textures That Fool Visual Object Tracking conference October 2019
Online Object Tracking: A Benchmark conference June 2013
Generating 3D Adversarial Point Clouds conference June 2019
Cooling-Shrinking Attack: Blinding the Tracker With Imperceptible Noises conference June 2020
Point Transformer conference October 2021
Box-Aware Feature Enhancement for Single Object Tracking on Point Clouds conference October 2021
Beyond 3D Siamese Tracking: A Motion-Centric Paradigm for 3D Single Object Tracking in Point Clouds conference June 2022
PTTR: Relational 3D Point Cloud Object Tracking with Transformer conference June 2022
LG-GAN: Label Guided Adversarial Network for Flexible Targeted Attack of Point Cloud Based Deep Networks conference June 2020

Similar Records

Related Subjects