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

The prominent progress in generative models has significantly improved thereality of generated faces, bringing serious concerns to society. Since recentGAN-generated faces are in high realism, the forgery traces have become moreimperceptible, increasing the forensics challenge. To combat GAN-generatedfaces, many countermeasures based on Convolutional Neural Networks (CNNs) havebeen spawned due to their strong learning ability. In this paper, we rethinkthis problem and explore a new approach based on forest models instead of CNNs.Specifically, we describe a simple and effective forest-based method set called{\em ForensicsForest Family} to detect GAN-generate faces. The proposedForensicsForest family is composed of three variants, which are {\emForensicsForest}, {\em Hybrid ForensicsForest} and {\em Divide-and-ConquerForensicsForest} respectively. ForenscisForest is a newly proposed Multi-scaleHierarchical Cascade Forest, which takes semantic, frequency and biologyfeatures as input, hierarchically cascades different levels of features forauthenticity prediction, and then employs a multi-scale ensemble scheme thatcan comprehensively consider different levels of information to improve theperformance further. Based on ForensicsForest, we develop HybridForensicsForest, an extended version that integrates the CNN layers intomodels, to further refine the effectiveness of augmented features. Moreover, toreduce the memory cost in training, we propose Divide-and-ConquerForensicsForest, which can construct a forest model using only a portion oftraining samplings. In the training stage, we train several candidate forestmodels using the subsets of training samples. Then a ForensicsForest isassembled by picking the suitable components from these candidate forestmodels...

ForensicsForest Family: A Series of Multi-scale Hierarchical Cascade Forests for Detecting GAN-generated Faces