Impact of Media Forensics and Deepfake in Society

Ehsan Nowroozi; Seyedsadra Seyedshoari; Mohammadreza Mohammadi; Alireza Jolfaei

doi:10.1007/978-3-031-10706-1_18

Impact of Media Forensics and Deepfake in Society

Ehsan Nowroozi, Seyedsadra Seyedshoari, Mohammadreza Mohammadi, Alireza Jolfaei

College of Science and Engineering

Research output: Chapter in Book/Report/Conference proceeding › Chapter › peer-review

16 Citations (Scopus)

Abstract

Technologies for producing and modifying content for multimedia have reached to the point that they could now deliver an extremely high degree of realism. Free software tools accessible on the Internet enable anyone with no special expertise to generate incredibly convincing fake images and videos. These technologies provide plenty of new opportunities in various sectors such as fine arts, marketing, film-making, and video games. They could also be employed to influence thoughts of the society in election campaigns, attempt crime, blackmail, or defame individuals. Thus, automated and effective methods of identifying fake media content are urgently needed. This chapter aims to provide an assessment of approaches for validating multimedia integrity as well as detecting modified images and videos. Deepfakes generated using deep learning (DL) techniques will be highlighted particularly, as well as recent data-driven forensics strategies to counter them. The results emphasize the shortcomings of present forensics techniques, the most crucial concerns, emerging obstacles, and future research possibilities.

Original language	English
Title of host publication	Breakthroughs in Digital Biometrics and Forensics
Editors	Kevin Daimi, Guillermo Francia III, Luis Hernández Encinas
Place of Publication	Switzerland
Publisher	Springer Nature Switzerland AG
Pages	387-410
Number of pages	24
ISBN (Electronic)	978-3-031-10706-1
ISBN (Print)	978-3-031-10705-4, 978-3-031-10708-5
DOIs	https://doi.org/10.1007/978-3-031-10706-1_18
Publication status	Published - 2022

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

SDG 16 Peace, Justice and Strong Institutions

Keywords

Adversarial multimedia forensics
Cybersecurity
Deepfakes
Image and video forensics
Machine and deep learning
Multimedia forensics

Access to Document

10.1007/978-3-031-10706-1_18Licence: In Copyright

https://rdcu.be/eaZd5Licence: In Copyright

Cite this

@inbook{d1804f74b56d4d68ae294d379266e6c1,

title = "Impact of Media Forensics and Deepfake in Society",

abstract = "Technologies for producing and modifying content for multimedia have reached to the point that they could now deliver an extremely high degree of realism. Free software tools accessible on the Internet enable anyone with no special expertise to generate incredibly convincing fake images and videos. These technologies provide plenty of new opportunities in various sectors such as fine arts, marketing, film-making, and video games. They could also be employed to influence thoughts of the society in election campaigns, attempt crime, blackmail, or defame individuals. Thus, automated and effective methods of identifying fake media content are urgently needed. This chapter aims to provide an assessment of approaches for validating multimedia integrity as well as detecting modified images and videos. Deepfakes generated using deep learning (DL) techniques will be highlighted particularly, as well as recent data-driven forensics strategies to counter them. The results emphasize the shortcomings of present forensics techniques, the most crucial concerns, emerging obstacles, and future research possibilities.",

keywords = "Adversarial multimedia forensics, Cybersecurity, Deepfakes, Image and video forensics, Machine and deep learning, Multimedia forensics",

author = "Ehsan Nowroozi and Seyedsadra Seyedshoari and Mohammadreza Mohammadi and Alireza Jolfaei",

year = "2022",

doi = "10.1007/978-3-031-10706-1_18",

language = "English",

isbn = "978-3-031-10705-4",

pages = "387--410",

editor = "Kevin Daimi and {Francia III}, Guillermo and Encinas, {Luis Hern{\'a}ndez}",

booktitle = "Breakthroughs in Digital Biometrics and Forensics",

publisher = "Springer Nature Switzerland AG",

address = "Switzerland",

}

Impact of Media Forensics and Deepfake in Society. / Nowroozi, Ehsan; Seyedshoari, Seyedsadra; Mohammadi, Mohammadreza et al.
Breakthroughs in Digital Biometrics and Forensics. ed. / Kevin Daimi; Guillermo Francia III; Luis Hernández Encinas. Switzerland: Springer Nature Switzerland AG, 2022. p. 387-410.

Research output: Chapter in Book/Report/Conference proceeding › Chapter › peer-review

TY - CHAP

T1 - Impact of Media Forensics and Deepfake in Society

AU - Nowroozi, Ehsan

AU - Seyedshoari, Seyedsadra

AU - Mohammadi, Mohammadreza

AU - Jolfaei, Alireza

PY - 2022

Y1 - 2022

N2 - Technologies for producing and modifying content for multimedia have reached to the point that they could now deliver an extremely high degree of realism. Free software tools accessible on the Internet enable anyone with no special expertise to generate incredibly convincing fake images and videos. These technologies provide plenty of new opportunities in various sectors such as fine arts, marketing, film-making, and video games. They could also be employed to influence thoughts of the society in election campaigns, attempt crime, blackmail, or defame individuals. Thus, automated and effective methods of identifying fake media content are urgently needed. This chapter aims to provide an assessment of approaches for validating multimedia integrity as well as detecting modified images and videos. Deepfakes generated using deep learning (DL) techniques will be highlighted particularly, as well as recent data-driven forensics strategies to counter them. The results emphasize the shortcomings of present forensics techniques, the most crucial concerns, emerging obstacles, and future research possibilities.

AB - Technologies for producing and modifying content for multimedia have reached to the point that they could now deliver an extremely high degree of realism. Free software tools accessible on the Internet enable anyone with no special expertise to generate incredibly convincing fake images and videos. These technologies provide plenty of new opportunities in various sectors such as fine arts, marketing, film-making, and video games. They could also be employed to influence thoughts of the society in election campaigns, attempt crime, blackmail, or defame individuals. Thus, automated and effective methods of identifying fake media content are urgently needed. This chapter aims to provide an assessment of approaches for validating multimedia integrity as well as detecting modified images and videos. Deepfakes generated using deep learning (DL) techniques will be highlighted particularly, as well as recent data-driven forensics strategies to counter them. The results emphasize the shortcomings of present forensics techniques, the most crucial concerns, emerging obstacles, and future research possibilities.

KW - Adversarial multimedia forensics

KW - Cybersecurity

KW - Deepfakes

KW - Image and video forensics

KW - Machine and deep learning

KW - Multimedia forensics

UR - http://www.scopus.com/inward/record.url?scp=85153845568&partnerID=8YFLogxK

U2 - 10.1007/978-3-031-10706-1_18

DO - 10.1007/978-3-031-10706-1_18

M3 - Chapter

AN - SCOPUS:85153845568

SN - 978-3-031-10705-4

SN - 978-3-031-10708-5

SP - 387

EP - 410

BT - Breakthroughs in Digital Biometrics and Forensics

A2 - Daimi, Kevin

A2 - Francia III, Guillermo

A2 - Encinas, Luis Hernández

PB - Springer Nature Switzerland AG

CY - Switzerland

ER -

Impact of Media Forensics and Deepfake in Society

Abstract

UN SDGs

Keywords

Access to Document

Other files and links

Fingerprint

Cite this