I predict that the first use of this will be in campaign ads, where the technology will be used to make psychopathic politicians look like non-psychopaths.
Perfect for the political technique of “muddying the waters.” It is now possibly to undermine the credibility of absolutely any evidence … or at least any evidence that can be transmitted to the public.
It can be used like Auto-Tune for visuals as well as vocals. Get your face scanned in detail, digitally manipulate the model to higher symmetry/clearer skin/younger appearance, and thereafter it can be overlaid on your real face in real time or in postproduction. It’s an inverted picture of Dorian Gray: your real face is aging all the time, but to everyone who sees you through recordings, you look just as good as you did at 23. Or better!
Just in time to give me an all-purpose alibi for all sorts of potentially embarrassing situations. Who says modern technology can’t meet our most pressing needs?
@5 interesting! Will people’s suspicion of the truth of what is shown on the screen rise commensurately with the ability to doctor and distort videographic material? I suspect not. A new age of propaganda.
So, if I understand this correctly, “real time” means that someone could hijack a press conference, showing the world a person saying things totally different from what s/he actually said?
That’s a scary prospect.
Guess it’s time to do that remake of Metropolis
Ted @1. Or it could be used for the inverse function, depending upon the agenda of the manipulator. Release it the day before the election, and by the time the malicious agenda is exposed it will be too late.
Or manufacture false evidence of a crime, and put your enemies behind bars. The possibilities are endless.
John M. Burt @15: “So, if I understand this correctly, “real time†means that someone could hijack a press conference, showing the world a person saying things totally different from what s/he actually said?”
Given there’s no audio in the process, no they couldn’t. And “real time” is a dubious claim when you read the paper and note things like:
“For a given transferred facial expression, we need to synthesize a realistic target mouth region. To this end, we retrieve and warp the best matching mouth image from the target actor sequence. We assume that sufficient mouth variation is available in the target video.”
The re-enactment might take place in run time, but it can’t be applied to a live video feed. It requires pre-analysis.
Ze K @9: no, this is not a simulacrum nor is it related to the concept in any particularly interesting way. Baudrillard’s point was a simulation, or more precisely a copy, of a thing that was better by some notable metric than the original. The canonical example is the Disney castle vs. Neuschwanstein, but there are lots of other examples across the USA.
{ 18 comments }
Ted Lemon 03.24.16 at 2:45 pm
I predict that the first use of this will be in campaign ads, where the technology will be used to make psychopathic politicians look like non-psychopaths.
otpup 03.24.16 at 3:03 pm
Disturbing.
Niall McAuley 03.24.16 at 3:22 pm
I have no comment, I just like to say “artificially shaped mouth regions”.
Jane 03.24.16 at 4:38 pm
First airbrushing now animated airbrushing, wonderful :(
Adam Hammond 03.24.16 at 5:01 pm
Perfect for the political technique of “muddying the waters.” It is now possibly to undermine the credibility of absolutely any evidence … or at least any evidence that can be transmitted to the public.
Matt 03.24.16 at 5:17 pm
It can be used like Auto-Tune for visuals as well as vocals. Get your face scanned in detail, digitally manipulate the model to higher symmetry/clearer skin/younger appearance, and thereafter it can be overlaid on your real face in real time or in postproduction. It’s an inverted picture of Dorian Gray: your real face is aging all the time, but to everyone who sees you through recordings, you look just as good as you did at 23. Or better!
Glen Tomkins 03.24.16 at 6:10 pm
Just in time to give me an all-purpose alibi for all sorts of potentially embarrassing situations. Who says modern technology can’t meet our most pressing needs?
Bloix 03.24.16 at 7:00 pm
Couple this with 360-degree virtual reality viewers – IT’S THE END OF THE REAL WORLD!!
Joseph Brenner 03.24.16 at 11:58 pm
Ah, at last, the David Bowie channel, where all parts are
performed by David Bowie. Hamlet, Hamlet’s Father, Ophelia,
Grave diggers, Yorick…
NomadUK 03.25.16 at 2:15 am
I’m sure the Impossible Missions Force could put this to good use.
Otto Fazomi 03.25.16 at 10:35 pm
Real time translations ala Star Trek!
RNB 03.26.16 at 8:54 am
@5 interesting! Will people’s suspicion of the truth of what is shown on the screen rise commensurately with the ability to doctor and distort videographic material? I suspect not. A new age of propaganda.
John M. Burt 03.26.16 at 7:07 pm
So, if I understand this correctly, “real time” means that someone could hijack a press conference, showing the world a person saying things totally different from what s/he actually said?
That’s a scary prospect.
Guess it’s time to do that remake of Metropolis
Omega Centauri 03.28.16 at 1:03 am
Ted @1. Or it could be used for the inverse function, depending upon the agenda of the manipulator. Release it the day before the election, and by the time the malicious agenda is exposed it will be too late.
Or manufacture false evidence of a crime, and put your enemies behind bars. The possibilities are endless.
Nick 03.29.16 at 10:55 am
John M. Burt @15: “So, if I understand this correctly, “real time†means that someone could hijack a press conference, showing the world a person saying things totally different from what s/he actually said?”
Given there’s no audio in the process, no they couldn’t. And “real time” is a dubious claim when you read the paper and note things like:
“For a given transferred facial expression, we need to synthesize a realistic target mouth region. To this end, we retrieve and warp the best matching mouth image from the target actor sequence. We assume that sufficient mouth variation is available in the target video.”
The re-enactment might take place in run time, but it can’t be applied to a live video feed. It requires pre-analysis.
Nick 03.29.16 at 10:57 am
(sorry, some unclosed italics there around run time…)
Paul Davis 03.31.16 at 5:40 am
Ze K @9: no, this is not a simulacrum nor is it related to the concept in any particularly interesting way. Baudrillard’s point was a simulation, or more precisely a copy, of a thing that was better by some notable metric than the original. The canonical example is the Disney castle vs. Neuschwanstein, but there are lots of other examples across the USA.
hix 04.01.16 at 12:04 am
Ehat metric would that be, Neuschwanstein is so much better
Comments on this entry are closed.