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

Significant advances in deep learning have obtained hallmark accuracy ratesfor various computer vision applications. However, advances in deep generativemodels have also led to the generation of very realistic fake content, alsoknown as deepfakes, causing a threat to privacy, democracy, and nationalsecurity. Most of the current deepfake detection methods are deemed as a binaryclassification problem in distinguishing authentic images or videos from fakeones using two-class convolutional neural networks (CNNs). These methods arebased on detecting visual artifacts, temporal or color inconsistencies producedby deep generative models. However, these methods require a large amount ofreal and fake data for model training and their performance drops significantlyin cross dataset evaluation with samples generated using advanced deepfakegeneration techniques. In this paper, we thoroughly evaluate the efficacy ofdeep face recognition in identifying deepfakes, using different loss functionsand deepfake generation techniques. Experimental investigations on challengingCeleb-DF and FaceForensics++ deepfake datasets suggest the efficacy of deepface recognition in identifying deepfakes over two-class CNNs and the ocularmodality. Reported results suggest a maximum Area Under Curve (AUC) of 0.98 andan Equal Error Rate (EER) of 7.1% in detecting deepfakes using face recognitionon the Celeb-DF dataset. This EER is lower by 16.6% compared to the EERobtained for the two-class CNN and the ocular modality on the Celeb-DF dataset.Further on the FaceForensics++ dataset, an AUC of 0.99 and EER of 2.04% wereobtained. The use of biometric facial recognition technology has the advantageof bypassing the need for a large amount of fake data for model training andobtaining better generalizability to evolving deepfake creation techniques.

An Experimental Evaluation on Deepfake Detection using Deep Face Recognition