Google has recently released an open-source database which consists 3,000 deepfake videos, made by using algorithms. Over a year, Google worked with various actors to make these videos in order to tussle against deepfakes.
The videos show the recordings of actors speaking, showing their expressions and performing humdrum tasks. Google then used various publicly available deepfake tools to manipulate the original recorded videos and created plenty of deepfakes to use them as a reference to battle with new deepfakes on the internet.
For those who don't know what deepfakes are - they are manipulated videos made with the help of Artificial intelligence. The deepfake technology collects the voices and clips of various celebs, alter a bit and swap their persona with some unknown.
Earlier, Facebook CEO Mark Zuckerberg himself became a victim to these deepfakes. A clip was made on him and released on social media which is made by using the AI software. Some clips even consists pornography too.
Well, many experts and analysts currently believe that, these deepfake clips might affect the upcoming US presidential elections.
However, as soon as the data released, Google started working closely with researchers with a ray of hope to develop a detection tool to alleviate these deepfake videos.
A blog post in Google AI blog about "Contributing data to deep fake detention research", said that: Google takes these issue on a serious note and shared some previous published AI principles. They also shared the works done behind this issue, everything about recording the clips of film actors.
The Google CEO also took to twitter sharing about the detecting deepfake challenges.
He stated, "Detecting deepfakes is one of the most important challenges ahead of us. Following our release of a synthetic audio dataset in Jan, we're releasing a large dataset of visual deepfakes to support researchers working on synthetic video detection #GoogleAI".
Google is not the only sole partner who's fighting for deepfakes, even Facebook and Microsoft is working closely with this problem. This month, Facebook had made an arrangement of 10milion dollars to deal with this issue. Their sole aim is to create a tool which equally aid the government, companies and media companies to identify the difference between deepfake and original.
While some still expostulate that, handling this in technical way will not solely work, it a group challenge to take down together, socially, politically and technically in various legal ways.
