Now Even You Can Realistically Impersonate Obama Using This Lip Syncing AI

Various people have predicted that artificial intelligence will soon take over our lives in more ways than we have ever imagined. AI robots are just a fraction of the humongous impact that artificial intelligence will have on our lives. We recently heard of an AI voice generator that can imitate nearly any voice and now researchers at the University of Washington have developed an AI that can produce fake video footage of the former US President Barack Obama.

The AI developed by the University of Washington team using an artificial neural network is so realistic that it matches the former president’s voice perfectly. It is becoming overly difficult to distinguish fake news from the real ones, and the Obama AI is just a crazy addition to our fear of being delivered the fake news very often. Photos were rejected as reliable proof a long time ago, and now, even videos will not be considered substantial evidence in many scenarios. Who are we going to trust then?

One of the researchers working on the project, Dr. Supasorn Suwajanakorn commented, “We developed an algorithm that can generate a believable video of Obama from his voice, based on a recurrent neural network that learns how to do this by analyzing hours of Obama’s weekly address footage. Unlike prior work, we never require the subject to be scanned or a speech database that consists of videos of many people saying predetermined sentences. We learn this from just existing footage. This has the potential to scale to anyone with minimal effort.”

The reason for picking Obama for building the AI impersonator was not a political bias but the fact that “deep learning requires lots of data and this data set is well-suited because it’s large: over 20 hours, easy to collect, contains only Obama in high-resolution, and public-domain, which is free for researchers to use.”

via Wired

The current research is essentially a sophistication of the technique that film-makers use with archival footage. They take some existing video and then edit it to the dialogue that they require. The AI impersonating Obama does the similar thing and makes the president’s mouth match the dialogue.

The AI is getting super intelligent, and while this may be exciting, it is incredibly scary as well. Meanwhile, you might see a ghost video on your TV, which was impersonated out of a deadly ghost from a scary movie.

What do you think about this advanced AI? Share your views in the comments section below.

Leave a Reply

Your email address will not be published. Required fields are marked *