Last night, BBC Newsnight highlighted new software for the video manipulation of facial expressions. Combined with advances in voice manipulation, the software has major implications for the credibility of video evidence.
Using a webcam, facial tracking software and video footage of a target person, Face2Face enables highly-convincing transfer of facial expressions, including distinct mouth and lip movements. In the video demonstration below, an actor manipulates the facial expressions of George Bush, Vladimir Putin and Donald Trump.
https://www.youtube.com/watch?v=ohmajJTcpNk
Unveiled early last year, the software was developed by researchers from Stanford University, the University of Erlangen-Nuremberg and the Max Planck Institute for Informatics. Although facial capture and re-mapping is not a new development, the Face2Face software greatly reduces the skills and resources required to achieve the effect.
The developers believe it "will pave the way for many new and exciting applications in the fields of VR/AR, teleconferencing, or on-the- fly dubbing of videos with translated audio" (Thies, J. et al 2016). Despite the impressive technical achievement, the potential for misuse is considerable, particularly in a political climate of “fake news" and White House approved “alternative facts”.
Last November, Adobe demonstrated early results from Project VoCo; an audio editing application that can create speech by replicating a speaker’s recorded voice. As Abobe describe it, “you can simply type in the word or words that you would like to change or insert into the voiceover. The algorithm does the rest and makes it sound like the original speaker said those words.”
The software, which is not currently planned for public release, raises serious issues for voice security, not least for banks and business using voice recognition for identification.
https://www.youtube.com/watch?v=I3l4XLZ59iw
Subscribe to FuJo's newsletter
