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

Recent studies have shown that the performance of forgery detection can beimproved with diverse and challenging Deepfakes datasets. However, due to thelack of Deepfakes datasets with large variance in appearance, which can behardly produced by recent identity swapping methods, the detection algorithmmay fail in this situation. In this work, we provide a new identity swappingalgorithm with large differences in appearance for face forgery detection. Theappearance gaps mainly arise from the large discrepancies in illuminations andskin colors that widely exist in real-world scenarios. However, due to thedifficulties of modeling the complex appearance mapping, it is challenging totransfer fine-grained appearances adaptively while preserving identity traits.This paper formulates appearance mapping as an optimal transport problem andproposes an Appearance Optimal Transport model (AOT) to formulate it in bothlatent and pixel space. Specifically, a relighting generator is designed tosimulate the optimal transport plan. It is solved via minimizing theWasserstein distance of the learned features in the latent space, enablingbetter performance and less computation than conventional optimization. Tofurther refine the solution of the optimal transport plan, we develop asegmentation game to minimize the Wasserstein distance in the pixel space. Adiscriminator is introduced to distinguish the fake parts from a mix of realand fake image patches. Extensive experiments reveal that the superiority ofour method when compared with state-of-the-art methods and the ability of ourgenerated data to improve the performance of face forgery detection.

AOT: Appearance Optimal Transport Based Identity Swapping for Forgery Detection