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

With the development of deep learning technology, the facial manipulationsystem has become powerful and easy to use. Such systems can modify theattributes of the given facial images, such as hair color, gender, and age.Malicious applications of such systems pose a serious threat to individuals'privacy and reputation. Existing studies have proposed various approaches toprotect images against facial manipulations. Passive defense methods aim todetect whether the face is real or fake, which works for posterior forensicsbut can not prevent malicious manipulation. Initiative defense methods protectimages upfront by injecting adversarial perturbations into images to disruptfacial manipulation systems but can not identify whether the image is fake. Toaddress the limitation of existing methods, we propose a novel two-tierprotection method named Information-containing Adversarial Perturbation (IAP),which provides more comprehensive protection for {facial images}. We use anencoder to map a facial image and its identity message to a cross-modeladversarial example which can disrupt multiple facial manipulation systems toachieve initiative protection. Recovering the message in adversarial exampleswith a decoder serves passive protection, contributing to provenance trackingand fake image detection. We introduce a feature-level correlation measurementthat is more suitable to measure the difference between the facial images thanthe commonly used mean squared error. Moreover, we propose a spectral diffusionmethod to spread messages to different frequency channels, thereby improvingthe robustness of the message against facial manipulation. Extensiveexperimental results demonstrate that our proposed IAP can recover the messagesfrom the adversarial examples with high average accuracy and effectivelydisrupt the facial manipulation systems.

Information-containing Adversarial Perturbation for Combating Facial Manipulation Systems