arrow
返回

Context-Guided Black-Box Attack for Visual Tracking

delete2024-01-01
delete0
PRE
AI
X
Xingsen Huang
D
Deshui Miao
H
Hongpeng Wang *
王
王耀威 (Yaowei Wang)
X
Xin Li *
DOI:10.1109/TMM.2024.3382473delete
delete原文链接
delete原文求助
delete分享
delete收藏
摘要

摘要

En 中文
With the recent advancement of deep neural networks, visual tracking has achieved substantial progress in tracking accuracy. However, the robustness and security of tracking methods developed based on current deep models have not been thoroughly explored, a critical consideration for real-world applications. In this study, we propose a context-guided black-box attack method to investigate the robustness of recent advanced deep trackers against spatial and temporal interference. For spatial interference, the proposed algorithm generates adversarial target samples by mixing the information of the target object and the similar background regions around it in an embedded feature space of an encoder-decoder model, which evaluates the ability of trackers to handle background distractors. For temporal interference, we use the target state in the previous frame to generate the adversarial sample, which easily fools the trackers that rely too heavily on tracking prior assumptions, such as that the appearance changes and movements of a video target object are small between two consecutive frames. We assess the proposed attack method under both CNN-based and transformer-based tracking frameworks on four diverse datasets: OTB100, VOT2018, GOT-10 k, and LaSOT. The experimental results demonstrate that our approach substantially deteriorates the performance of all these deep trackers across numerous datasets, even in the black-box attack mode. This reveals the weak robustness of recent deep tracking methods against background distractors and prior dependencies.
Keyword:
Target tracking
Feature extraction
Visualization
Transformers
Interference
Image reconstruction
Robustness
Visual tracking
adversarial attack

期刊

IEEE Transactions on Multimedia 封面图
IEEE Transactions on Multimedia
IF:
9.7
论文数:
4.5K
被引数:
2.4W

机构

H
harbin institute of technology
学者数:
8.0W
论文数: 6.6W
被引数: 66
P
Peng Cheng Laboratory
学者数:
1.7K
论文数: 1.8K
被引数: 2.0K
引用论文

引用论文

Transcriptome analysis of growth heterosis in pearl oyster Pinctada fucata martensii
err2018-10-12
err0
errOAAI
errJingmiao Yang; Shaojie Luo; Junhui Li; Zhe Zheng; Xiaodong Du; Yuewen Deng
err分享
err收藏
err分享
err收藏
Pasadena: Perceptually Aware and Stealthy Adversarial Denoise Attack
err2022-01-01
err13
PREAI
errCheng, Yupeng; Guo, Qing; Juefei-Xu, Felix; Lin, Shang-Wei; Feng, Wei; Lin, Weisi; Liu, Yang
err分享
err收藏
Validation of the Environmental Audit Tool in both purpose‐built and non‐purpose‐built dementia care settings
err2011-08-07
err0
errOAAI
errRonald Smith; Richard Fleming; Lynn Chenoweth; Yun‐Hee Jeon; Jane Stein‐Parbury; Henry Brodaty
err分享
err收藏
学者 查看更多内容