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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Despite the fact that DeepFake forgery detection algorithms have achievedimpressive performance on known manipulations, they often face disastrousperformance degradation when generalized to an unseen manipulation. Some recentworks show improvement in generalization but rely on features fragile to imagedistortions such as compression. To this end, we propose Diff-ID, a concise andeffective approach that explains and measures the identity loss induced byfacial manipulations. When testing on an image of a specific person, Diff-IDutilizes an authentic image of that person as a reference and aligns them tothe same identity-insensitive attribute feature space by applying aface-swapping generator. We then visualize the identity loss between the testand the reference image from the image differences of the aligned pairs, anddesign a custom metric to quantify the identity loss. The metric is then provedto be effective in distinguishing the forgery images from the real ones.Extensive experiments show that our approach achieves high detectionperformance on DeepFake images and state-of-the-art generalization ability tounknown forgery methods, while also being robust to image distortions.

Diff-ID: An Explainable Identity Difference Quantification Framework for DeepFake Detection