A New and Threatening Era of Digital Image

Camera is one of the most impressive technologies of mankind. This technology has come a long way from capturing black and white pictures to HD and 3D images. And in no time this technology is accessible to everyone. Nowadays everyone has a smartphone with a camera in it.

The digital camera applications in our cell phones have become increasingly innovative with time. Capturing high-definition pictures can do many other things. Like increasing brightness, removing pimples or my favorite adding a dog filter on your face. But the technology has reached a point where it can be misused like swapping your face image with someone else and making it look true.

There are many photoshop tools available on the internet which can edit your images any way you want. You might take a picture in your hometown and edit it to make sure you are on a holiday in Paris. You can crop your images and put them in a movie and make yourself a movie star. Your imagination is the only limit.

However these tools might seem fun and creative, there are also some dark stories related to them.

You might not be familiar with the term “synthetic media”, but you have somewhere read the word “Deepfake”. They both mean the same thing. But what do they mean and why are they so dangerous?

What is Deepfake?

Deepfake, let’s divide this word into “deep” and “fake”. It is a mixture of machine learning or deep learning with something that is not real. It is an artificial intelligence technology that puts fake images and voices together.

It is a digital representation of manipulative videos and images created by artificial intelligence algorithms in such a way that it appears to be real.

Let’s talk in simple terms.

Have you ever seen the above image? It’s where a famous Hollywood actor, Tom Cruise, is making Tiktok videos and showing magic tricks. The magic trick might be amazing but what the real magic was, is that it is not actually Tom Cruise.

If you thought that this is Tom Cruise then you are not alone. Many people believe that this is their beloved star. The quality of technology is so convincing that it produces hyper-realistic results which make us believe that it happened for real. This is what deepfake does, it makes people do or say things that they have never said or done.

How did it come into existence?

In 2017, a reddit user named “Deepfake” shared deepfake videos. The videos contain celebrity images swapped with the actresses in pornographic videos. Some other videos contain actors’ faces swapped with the faces of another actor in another movie. This was where the term “Deepfake” came into existence.

Uses of Deepfakes

Many filmmakers use this technology for special effects and animation purposes. After the tragic death of the Fast and Furious actor, Pual Walker, in a car accident. The filming team managed to complete the film without the actor using this technology. It took the studio months to produce a convincing deepfake.

It might have some positive uses but the majority of its uses consist of defamation and spreading fake news. The technology is misused to make scams and hoaxes. Many celebrities and females have become victims of pornographic deepfakes. Identity fraud and theft had been committed using this technology.

You may receive a call from your boss asking you to send him classified information. Maybe that call was not from your boss and the voice was exactly the same. Deepfake is not just limited to image but voice also. You can also receive some documents which may look authentic but are not. It is important for an organization to verify documents it receives from employees and customers.

Deepfake and Fake news

Deepfake is might be the most indestructible weapon to spread fake news. It can harm any individual reputation. The technology can be used by the opponent or supporter of the opponent in the election to produce manipulative content.

A political career can be destroyed by making a political candidate appear and do things he has never done. Such a type of video can spread fake news which can alter the decision of the public or even can land the participant in great trouble.

How to create a Deepfake

There are many different ways of creating a deepfake. Artificial intelligence deep-learning algorithms study the image or video of a target person from different angles to mimic the behavior and speech. A system known as GAN, Generative Adversarial Network, is used to create fake images. The system uses two sets of algorithms that learn the characteristics of an image and produce a fake one.

Other methods include capturing and storing numerous images of two people for swapping. The artificial intelligence coding identifies the similarity between two faces and swaps them.

Creating a deep fake video may require thousands of different pictures which is why celebrities are common targets of this. But the technology is still evolving and getting better. It is possible that in the future it may require a few images to create a convincing false video.

Identifying a Deepfake

A manipulative video can fool many people with its convincing quality, but not all deepfake are that good. There are some poor deepfake that are easy to spot. And a good deepfake is still not that accurate. All it needs is a good observation to identify it. Here are certain things you need to view to identify a deepfake

•  Movement of eyes

•  Skin Color

•  Unnatural Face Expressions

•  No or Fewer Eyes Blinking

•  Lip Sync with Voice

If it is the artificial intelligence that has created deep fakes then it is artificial intelligence that can help us identify a deepfake. Many companies have started working on artificial intelligence systems which can identify if the image or video is being deepfake.

Things to keep in Mind

Keeping the above-mentioned points in mind, you might be able to identify a fake video. But the quality of these videos can be very convincing which is why you should not believe and share everything that appears on the internet. Make sure you check if the source of the news is authentic.

Author Bio

James Watson is a contributor and technical content writer. He has written some articles for the readers which include Artificial Intelligence, Machine Learning, Cyber Security, Fintech, Crypto, Finance and Business, etc.