Publication: EfficientNets for DeepFake Detection: Comparison of Pretrained Models
Дата
2021
Авторы
Journal Title
Journal ISSN
Volume Title
Издатель
Аннотация
© 2021 IEEE.Rapid advances in media generation techniques have made the creation of AI-generated fake face videos more accessible than ever before. In order to accelerate the development of new ways to expose forged videos, Facebook created Deep Fake Detection Challenge (DFDC), which demonstrated multiple approaches to solve this problem. Analysis of top-performing solutions revealed that all winners used pre-trained EfficientNet networks, which was finetuned on videos containing face manipulations. Because of this observation, we decide to compare the performance of EfficientNets models within the task of detecting fake videos. For comparison, we use models, based on the highest-performing entrant of DFDC, entered by Selim Seferbekov, and the DFDC dataset as training data. Our experiments show that there is no strong correlation between model performance and its size. The best accuracy was achieved by B4 and B5 models.
Описание
Ключевые слова
Цитирование
Pokroy, A. A. EfficientNets for DeepFake Detection: Comparison of Pretrained Models / Pokroy, A.A., Egorov, A.D. // Proceedings of the 2021 IEEE Conference of Russian Young Researchers in Electrical and Electronic Engineering, ElConRus 2021. - 2021. - P. 598-600. - 10.1109/ElConRus51938.2021.9396092
URI
https://www.doi.org/10.1109/ElConRus51938.2021.9396092
https://www.scopus.com/record/display.uri?eid=2-s2.0-85104792685&origin=resultslist
http://gateway.webofknowledge.com/gateway/Gateway.cgi?GWVersion=2&SrcAuth=Alerting&SrcApp=Alerting&DestApp=WOS_CPL&DestLinkType=FullRecord&UT=WOS:000669709800132
https://openrepository.mephi.ru/handle/123456789/23975
https://www.scopus.com/record/display.uri?eid=2-s2.0-85104792685&origin=resultslist
http://gateway.webofknowledge.com/gateway/Gateway.cgi?GWVersion=2&SrcAuth=Alerting&SrcApp=Alerting&DestApp=WOS_CPL&DestLinkType=FullRecord&UT=WOS:000669709800132
https://openrepository.mephi.ru/handle/123456789/23975