Some commenters called up the possibility of abuse in adding words not originally found in the audio file. What if it were used to trick people or to abuse known personalities or opponents? Simply put, as one classic comment came in, "What could possibly go wrong?"
The point was not lost on Tyler Lee. He said that "from an ethical standpoint we have to wonder if such a software should ever be released to the masses."
[ ... ]
It will work if you feed Project VoCo about 20 minutes of audio featuring that someone's voice. Then it's good to go. The software proceeds to let you edit that speech. No Film School said "all you need is around 20 minutes of recorded speech for the algorithm to kick into gear for replication purposes. It analyzes the speech, breaks it down into phonemes, transcribes it, and creates the voice model."
http://ift.tt/2fpX64Y
The point was not lost on Tyler Lee. He said that "from an ethical standpoint we have to wonder if such a software should ever be released to the masses."
[ ... ]
It will work if you feed Project VoCo about 20 minutes of audio featuring that someone's voice. Then it's good to go. The software proceeds to let you edit that speech. No Film School said "all you need is around 20 minutes of recorded speech for the algorithm to kick into gear for replication purposes. It analyzes the speech, breaks it down into phonemes, transcribes it, and creates the voice model."
http://ift.tt/2fpX64Y
Video speech not sine qua non of authenticity. Adobe VoCo fauxtoshops audio track
0 commentaires:
Enregistrer un commentaire