DEEP FAKES
Technology is advancing day by day which has lead to invention of high quality fake videos with the help of your computer within hours. This technology is called DEEP FAKES. Now it is very easy to make synthetic videos. Deep Fakes uses Neural Networks to make fake videos. If you have ever read Artificial Intelligence you would have known Neural networks. Basically Neural Networks are those computer systems whose basic design is inspired by the human brain.In normal computer system consists of input and output. Functions are used to convert input into output after processing the information provided in the input. Programmer exactly defines the work to be performed by the particular function.But in case of Neural networks the functions are not designed by programmers but the function formation takes place on its own. So Neural Networks Function is hidden. These hidden functions are trained like humans with different examples to perform various work. Actually Neural networks creates its own function to perform given work.
Neural Networks are shown input and output multiple times, so that the function could be designed by it without any human intervention.For example a word "Hello" is written on the page with computer default fonts , today most computer system is designed to read the printed text , so normal computers can easily read the text "Hello" written on a page with computer default fonts.
If normal computer systems are given work to read a human's language they cannot read the human written text.
But Neural networks can read even human written text without difficulty if they are trained to read it.
So Neural networks must be trained to read Hello written by 20 persons and understand it points to Hello. We can increase the accuracy of Neural networks by increasing the number of examples text provided by us.So Neural networks will be able to generate the function by itself. Now Neural networks can easily understand the text or work.
USE OF NEURAL NETWORKS
Neural Networks are widely used by Snapchat and Instagram filters. Snapchat Neural Networks are trained to analyse each person's face and apply goggles or any other equipment to your face. Their Neural networks are shown a number of faces because every one's nose and eyes may not be at same position.After showing the faces, pointers are marked on each face so that Neural Networks can understand and identify the position of Nose, eyes, mouth and analyse the face.
Snapchat bought a company named Looksery because this company Looksery was expertised to use Neural networks for face recognition.
DEEP FAKES USED FOR FAKE OBAMA VIDEO
Deep Fakes was used to create a fake Obama's video, various footages of Obama was given to Neural networks to understand the body language , style of speaking, and how the voice of obama changes during speaking. So Neural networks can make multiple videos, as Obama also has different ways of speaking. This technology can be used in making movies. Now expenditure of movies can be reduced to an extent.What do you think about its uses?
USE OF DEEP FAKES IN INDIA
Deep fakes has positive as well as negative aspects. In India , where paid media ,IT cells, Fake news has flooded the country. In our country where police even not able to identify fake pro pakistan slogans video in Bihar, how will they understand the ultra HD videos of Deep Fakes.
Deep fakes are not threat to common people, instead journalist , celebrities , politicians fake videos can be made easily.
I want to make aware you about this Deep fakes technology.
SOLUTIONS TO DEEP FAKES
First solution is "Tit for Tat", if a Neural network can be trained to make fake videos. Truly Neural networks can be trained to detect fake videos probably Anti Fake Neural networks.
Second solution is store original verified videos inside Blockchain.So that we are aware of Real videos and fake videos.If you know bitcoin you know Blockchain.Blockchain can store history and it cannot be tampered.
This is dummy text. It is not meant to be read. Accordingly, it is difficult to figure out when to end it. But then, this is dummy text. It is not meant to be read. Period.