Headline, February 10 2019/ '' ' DEEP FAKES DEAN ' '' : DISINFORMATION

'' ' DEEP FAKES DEAN ' '' : 


REMEMBER : ON THE WORLD STUDENTS SOCIETY the fight against deepfakes and other forms of online disinformation will require nearly constant reinvention.

DEEPFAKE TECHNOLOGY HAS IMPROVED at a rate that surprises A.I. experts, and there is little reason to believe it will slow.

Deepfakes should benefit from one of the few tech industry axioms that have held up over the years. Computers always get more powerful, and there is always more data.

That makes the so called machine-learning software that helps create deepfakes more effective.

''It is getting easier, and it will continue to get easier. There is no doubt about it,'' said Matthias Niessner, a professor of computer science at the Technical University of Munich who is working with Google on its deepfake research. ''That trend will continue for years''.

The question is : Which side will improve more quickly? Researchers like Dr. Niessner are working to build systems that can automatically identify and remove deepfakes.

Google recently said that any academic or corporate researcher could download its collection of  synthetic videos and use them to build tools for identifying deepfakes. Like deepfake creators,  deepfake detectors learn their skills by analyzing images.

Detectors can also improve by leaps and bounds. But that requires a constant stream of new data representing the latest deepfake techniques around the Internet, Dr. Niessner and other researchers said.

The videos collection is essentially a syllabus of digital trickery for computers. By analyzing all of those images, A.I. systems learn how to watch for fakes. Facebook recently did something similar, using actors to build fake videos and then releasing them to outside researchers.

Engineers at a Canadian company called Dessa, which specializes in artificial intelligence, recently tested a deep fake detector that was built using Google's synthetic videos.

It could identify the Google's videos with almost perfect accuracy. But when they tested their detector on deepfake videos plucked from across the Internet, it failed more than 60 percent of the time.

They eventually fixed the problem, but only after rebuilding their detector with help from videos found ''in the wild,'' not created with paid actors - proving that a detector is only as good as the data used to train it.

Their tests showed that the fight against deepfakes and other forms of online disinformation will require nearly constant reinvention.

Several hundred synthetic videos are not enough to solve the problem, because they don't necessarily share the characteristics of fake videos being distributed today, much less in the years to come.

''Unlike other problems, this one is constantly changing,'' said Ragavan Thurairatnam, Dessa's founder and head of machine learning.

Though activists and artists release occasionally release deepfakes as a way of showing how these videos could shift the political discourse online, these techniques are not widely used to spread disinformation.

They are most used to spread humor or fake photography, according to Facebook, Google and others who track the progress of deepfakes.

Right now deepfake videos have subtle imperfections that can be readily detected by automate systems, if not by the naked eye. But some researchers argue that the improved technology will be powerful enough to create fake images without these tiny defects.

Companies like Google and Facebook hope they will have reliable detectors in place before that happens.

''In the short term, detectors will be reasonably effective,'' said Mr. Kambhampati, the Arizona State professor. ''In the longer term, I think it will be impossible to distinguish between the real pictures and the fake pictures.''

With respectful dedication to the Students, Professors and Teachers of the world. See Ya all prepare and register for Great Global Elections of The World Students Society : wssciw.blogspot.com and Twitter - !E-WOW! - The Ecosystem 2011:

''' Pressed Parameter '''

Good Night and God Bless

SAM Daily Times - the Voice of the Voiceless


Post a Comment

Grace A Comment!