We hope you can join us on THURSDAY, MARCH 2ND at 12:30PM CET for our
next 1-hour webinar with Fernando Pérez-González on “What's Up with
Image and Video Forensics!”

[6:30 a.m. New-York] – [12:30 p.m. Paris/Vigo] – [6:30 p.m. Beijing]

TO JOIN THE WEBINAR, PLEASE REGISTER TO RECEIVE MORE DETAILS ON HOW TO
CONNECT. The registration form can be found at:
https://forms.gle/9JCc6NBgM1x2kZK6A.
*If you wish to promote a EURASIP journal special issue, conference,
event, or new image/video database at an upcoming webinar, please reply
to this email with additional details.

TITLE: What's Up with Image and Video Forensics?

ABSTRACT: This talk will be a journey through more than 20 years of
research in video & image forensics, which was born to respond to
challenges of the digital era such as integrity and authenticity
verification and source attribution. We will examine some of the
characteristics that have made it possible, from the use of intrinsic
features in the acquisition devices, such as lens aberration or
demosaicing filters to the traces left by the postprocessing, in order
to grasp how methods have evolved from making use of relatively simple
statistical models to much more complex ones that rely on deep learning.
In particular, we will see this evolution with two classic problems: 1)
the identification of the device that captured a given image, to which
the use of the PRNU (Photo Response Non-Uniformity) opened the door and
which has currently expanded the type of fingerprints that can be
extracted and used successfully, and 2) the detection and localization
of image manipulations through the traces of double compression, which
now benefits from the ability of DNNs to learn models with richer
dependencies.
But not only forensic techniques have evolved: first, the popularization
of computational photography with lens distortion corrections,
high-dynamic range algorithms, video stabilization or AI-based filters,
etc.; second, the ever more prevalent use of social networks for image
and video exchange often with proprietary transcoding; and third, the
advent of Deep Fakes, imply that many of the techniques developed in the
past must be continuously revised to respond to all these novel and
demanding challenges.
Last, we will briefly discuss the ethical and legal implications of
image and video forensics, especially when it comes to the use of AI.

BIO: Fernando Pérez-González is a Professor with the School of
Telecommunication Engineering, University of Vigo, Spain, where he leads
the Signal Processing for Communications Group. In the late 90's, he did
pioneering work in the use of statistical and information theoretic
techniques for digital watermarking and later moved to multimedia
forensics and privacy-preserving signal processing. He has published
more than 250 papers in international journals and conference
proceedings and has co-authored 15 patent families. He has been
Principal Investigator of more than 45 contracts with industry.
Fawrensian, the commercial image forgery localization software he
developed with his team, is used worldwide to verify the integrity of
more than 1 million documents daily. He is currently involved in two
European projects: UNCOVER, that is pushing the state of the art in
steganalysis, and TRUMPET, which develops secure tools to minimize
privacy leakage in federated learning. He has been Senior Area Editor of
the IEEE Transactions on Information Forensics and Security, and Editor
in Chief of the EURASIP Journal on Information Security. Fernando is a
Fellow of the IEEE and of the AAIA, and member of the Royal Galician
Academy of Sciences.

Webinar videos are available online at
https://vimeo.com/showcase/8005816.