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

Deep neural networks (DNNs) are vulnerable to backdoor attacks which can hidebackdoor triggers in DNNs by poisoning training data. A backdoored modelbehaves normally on clean test images, yet consistently predicts a particulartarget class for any test examples that contain the trigger pattern. As such,backdoor attacks are hard to detect, and have raised severe security concernsin real-world applications. Thus far, backdoor research has mostly beenconducted in the image domain with image classification models. In this paper,we show that existing image backdoor attacks are far less effective on videos,and outline 4 strict conditions where existing attacks are likely to fail: 1)scenarios with more input dimensions (eg. videos), 2) scenarios with highresolution, 3) scenarios with a large number of classes and few examples perclass (a "sparse dataset"), and 4) attacks with access to correct labels (eg.clean-label attacks). We propose the use of a universal adversarial trigger asthe backdoor trigger to attack video recognition models, a situation wherebackdoor attacks are likely to be challenged by the above 4 strict conditions.We show on benchmark video datasets that our proposed backdoor attack canmanipulate state-of-the-art video models with high success rates by poisoningonly a small proportion of training data (without changing the labels). We alsoshow that our proposed backdoor attack is resistant to state-of-the-artbackdoor defense/detection methods, and can even be applied to improve imagebackdoor attacks. Our proposed video backdoor attack not only serves as astrong baseline for improving the robustness of video models, but also providesa new perspective for more understanding more powerful backdoor attacks.

Clean-Label Backdoor Attacks on Video Recognition Models