English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Visual object tracking (VOT) has been widely adopted in mission-criticalapplications, such as autonomous driving and intelligent surveillance systems.In current practice, third-party resources such as datasets, backbone networks,and training platforms are frequently used to train high-performance VOTmodels. Whilst these resources bring certain convenience, they also introducenew security threats into VOT models. In this paper, we reveal such a threatwhere an adversary can easily implant hidden backdoors into VOT models bytempering with the training process. Specifically, we propose a simple yeteffective few-shot backdoor attack (FSBA) that optimizes two lossesalternately: 1) a \emph{feature loss} defined in the hidden feature space, and2) the standard \emph{tracking loss}. We show that, once the backdoor isembedded into the target model by our FSBA, it can trick the model to losetrack of specific objects even when the \emph{trigger} only appears in one or afew frames. We examine our attack in both digital and physical-world settingsand show that it can significantly degrade the performance of state-of-the-artVOT trackers. We also show that our attack is resistant to potential defenses,highlighting the vulnerability of VOT models to potential backdoor attacks.

Few-Shot Backdoor Attacks on Visual Object Tracking