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

Residual-domain feature is very useful for Deepfake detection because itsuppresses irrelevant content features and preserves key manipulation traces.However, inappropriate residual prediction will bring side effects on detectionaccuracy. In addition, residual-domain features are easily affected by imageoperations such as compression. Most existing works exploit eitherspatial-domain features or residual-domain features, while neglecting that twotypes of features are mutually correlated. In this paper, we propose a guidedresiduals network, namely GRnet, which fuses spatial-domain and residual-domainfeatures in a mutually reinforcing way, to expose face images generated byDeepfake. Different from existing prediction based residual extraction methods,we propose a manipulation trace extractor (MTE) to directly remove the contentfeatures and preserve manipulation traces. MTE is a fine-grained method thatcan avoid the potential bias caused by inappropriate prediction. Moreover, anattention fusion mechanism (AFM) is designed to selectively emphasize featurechannel maps and adaptively allocate the weights for two streams. Theexperimental results show that the proposed GRnet achieves better performancesthan the state-of-the-art works on four public fake face datasets includingHFF, FaceForensics++, DFDC and Celeb-DF. Especially, GRnet achieves an averageaccuracy of 97.72% on the HFF dataset, which is at least 5.25% higher than theexisting works.

Exposing Deepfake Face Forgeries with Guided Residuals