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

While DeepFake applications are becoming popular in recent years, theirabuses pose a serious privacy threat. Unfortunately, most related detectionalgorithms to mitigate the abuse issues are inherently vulnerable toadversarial attacks because they are built atop DNN-based classificationmodels, and the literature has demonstrated that they could be bypassed byintroducing pixel-level perturbations. Though corresponding mitigation has beenproposed, we have identified a new attribute-variation-based adversarial attack(AVA) that perturbs the latent space via a combination of Gaussian prior andsemantic discriminator to bypass such mitigation. It perturbs the semantics inthe attribute space of DeepFake images, which are inconspicuous to human beings(e.g., mouth open) but can result in substantial differences in DeepFakedetection. We evaluate our proposed AVA attack on nine state-of-the-artDeepFake detection algorithms and applications. The empirical resultsdemonstrate that AVA attack defeats the state-of-the-art black box attacksagainst DeepFake detectors and achieves more than a 95% success rate on twocommercial DeepFake detectors. Moreover, our human study indicates thatAVA-generated DeepFake images are often imperceptible to humans, which presentshuge security and privacy concerns.

AVA: Inconspicuous Attribute Variation-based Adversarial Attack bypassing DeepFake Detection