A Novel Machine Learning based Method for Deepfake Video Detection in Social Media

Alakananda Mitra, Saraju P. Mohanty, Peter Corcoran, Elias Kougianos

Research output: Chapter in Book or Conference Publication/ProceedingConference Publicationpeer-review

40 Citations (Scopus)

Abstract

With the advent of deepfake videos, video forgery has become a serious threat. Videos in social media are the most common and serious targets. There are some existing works for detecting deepfake videos but very few attempts have been made for videos in social media. This paper presents a neural network based method to detect fake videos. A model, consisting of a convolutional neural network (CNN) and a classifier network is proposed. Three different structures, XceptionNet, InceptionV3 and Resnet50 have been considered as the CNN modules and a comparative study has been made. Xception Net has been chosen in the proposed model and paired with the proposed classifier for classification. We used the FaceForensics++ dataset to reach the best model. Our model integrated in the algorithm detects compressed videos in social media.

Original languageEnglish
Title of host publicationProceedings - 2020 6th IEEE International Symposium on Smart Electronic Systems, iSES 2020
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages91-96
Number of pages6
ISBN (Electronic)9780738142647
DOIs
Publication statusPublished - Dec 2020
Event6th IEEE International Symposium on Smart Electronic Systems, iSES 2020 - Virtual, Chennai, India
Duration: 14 Dec 202016 Dec 2020

Publication series

NameProceedings - 2020 6th IEEE International Symposium on Smart Electronic Systems, iSES 2020

Conference

Conference6th IEEE International Symposium on Smart Electronic Systems, iSES 2020
Country/TerritoryIndia
CityVirtual, Chennai
Period14/12/2016/12/20

Keywords

  • Compressed Video
  • Convolutional Neural Network (CNN)
  • Deep Learning
  • Deepfake
  • Depthwise Separable Convolution
  • Social Media
  • Transfer Learning

Fingerprint

Dive into the research topics of 'A Novel Machine Learning based Method for Deepfake Video Detection in Social Media'. Together they form a unique fingerprint.

Cite this