Adobe’s New Audio Tool VoCo Can Change What Anyone Says By Editing Speech, raises privacy concerns
How would you like to take someone’s speech available online and put your words into it? Till now this was impossible but now Adobe has been testing a software which could let you do just that.
Adobe has demonstrated its latest software — codenamed Project VoCo — that’s capable of easily editing the speech of people by simply cutting, pasting, and adding words to a digital transcript.
Adobe VoCo lets you edit recorded speech so that you can alter what that person said or create an entirely new sentence using their voice. Like its masterful Adobe Photoshop product which changed the way we view images and documents, the VoCo “photoshop but for audio.” may change the way we edit our speech.
Adobe VoCo (voice conversion), presents the user with a text box which shows the audio content the user wants to edit. Using the onscreen tools, users can move or swap words around, delete fragments, or type in entirely new words they want the target speaker to speak. When you type in a new word, VoCo deconstructs the text into code and reconstructs into speech matching the speaker voice. Adobe Voco then lets you play and listen to the new clip.
VoCo process large amount of voice data, about 20 minutes right now, and breaks it down into phonemes (each of the distinct sounds that make up a spoken language), and then attempting to create a voice model of the speaker—presumably stuff like cadence, stresses, quirks, etc.,
The software is still under development by Adobe Research and a team from Princeton and Adobe has yet to provide a release date or the cost of buying the software.
At a live demo in San Diego on Thursday, Adobe took a digitized recording of a man saying “and I kissed my dogs and my wife” and changed it to say “and I kissed Jordan three times”.
The edit took seconds and simply involved the operator overtyping a transcript of the speech and then pressing a button to create the synthesized voice track.
The applications of Adobe VoCo are huge in fields like movies and ads where the makers don’t have to make the actors speak again and again. But on the other hand, the misuse risks of Adobe VoCo is also huge. In an already charged atmosphere around the world, anti-social elements can use it to deconstruct any legitimate speech and edit it to slander and malign the personality. The software can also be used to alter news in such a way to create controversy and sensationalism by the TRP hungry mainstream media.
Then there is always the risk of terrorists using the tool to reprogram voice of prominent leaders and cause terror or bring down markets. Imagine using this tool, a terrorist or criminal edits President Obama’s sound byte to make him say that he is closing down Nasdaq. This edited speech could cause a selloff around the world. It remains to be seen whether Adobe introduces some protection to its tool to make it difficult for cyber criminals and terrorists to use it.