Deepfake AI can now make singing and a talking head using just a photo, by adding an audio clip to it.
Good news? Or do you think it is creepy too?
Well, it is not time to disclose opinions as the process has already gone on hands and has also gone on YouTube. Singing Einstein and other vocal political leaders have taken over YT.
Samsung’s AI research center and researchers from Imperial College in London have made a deep fake AI algorithm that generates a talking or a singing video using nothing but a person’s picture, no matter what age it mattered to and an audio clip. Like earlier, researchers have utilized machine learning algorithms to generate these deepfakes.
While some researchers look at this as a great way to progressing towards the future, a lot of people think of this as a great way to manipulate images to make fake videos. This kind of technology is already popular in the united states, as a lot of ministers and political men have faced false accusations due to online videos made using deepfake.
The audience has already seen creepy videos of Mona Lisa and other dead famous people coming alive in videos and they think of this as a step forward toward making the internet a scarier place for a lot of normal people.
The researchers, however, faced the fact that the general audience does not buy fake videos. The audience is too clever. In fact, when the researchers tried these videos on 25 people during testing, they have found that 12 of them have recognized the alters.
However, scientists hope that they could take better footsteps towards creating better footsteps in the future to making realistic videos.