Deepfakes-A blend of Deep Learning

In this post,I am going to explain about deep fake which is an artificial human based human image synthesis technique and the danger that it possesses. It can be used to superimpose impose image over videos. Deep fakes uses an artificial neural network and the power of the graphics processor and three to four gigabytes of storage space to generate the fake video. The limitation of deepfake videos are that it requires a lot of visual material from the person to be inserted in order to learn which image aspects have to be exchanged, using the previously mentioned deep learning algorithm based on the video sequences and images.

The above images are generated by using a deep fake software in which different output videos are generated according to the inputs you set.Now we can talk about disadvantages that it posses-:

It's blurring the line of reality-Deep fake videos are impossible to decrypt and presently no software has been made to counter this problem.It is becoming very difficult to differentiate between fake and real one.

Deepfakes in pornography-Deepfakes has been actively used in pornography videos of celebrity and many videos of celebrities have been surfaced like Gal Gadot, Emma Watson, Katy Perry, Taylor Swift or Scarlett Johansson.The scenes were not real and they have been generated using artificial intelligence. The first deepfake that captured world's attention was of Daisy Ridely. After that more and more videos have been surfacing.

Deepfakes in politics-It is one of the most serious issue faced by deepfake. Individual agendas can be fulfilled by using politicians name as well as videos.Deep fakes have already been used to misinterpret politicians.In countries like India it can cause havoc where various cases of mob lynching have been found by fake whatsapp news.It can cause war between big countries if videos of politicians like Kim Jong-un or Trump is misinterpreted or in worst case it can cause world war.