From Bruce Willis deepfake movies where Barack Obama calls Donald Trump a "complete idiot" to porn videos starring your favorite stars that piss off feminists. I've created some amazingly terrifying videos just by collecting machine learning models. Deepfake is a technique that uses a powered computer and deep learning to create fake images or videos of someone impersonating another person. Before delving deeper into the chaotic world of deepfakes, we must first understand their history and how they work.
Deepfake first came to prominence when his Reddit user named "Deepfakes" developed and published a machine learning algorithm that perfectly transformed celebrity faces into pornographic videos of him. The idea of video manipulation itself is not new. In the 1990s, some universities conducted academic research on computer vision. At the time, video manipulation was focused on AI and ML, just changing videos of people talking and merging them with another audio his track. The reason there are more celebrities and public figures than deepfake victims is that there is so much data, such as images and videos, that models can be trained to better cover their physical characteristics.
Deepfake works with his two different machine learning models that implement his GAN, a generative adversarial network. A GAN has two neural networks that challenge each other to create products from previous data. The first component is called a generator. Create images or videos from datasets using thought knowledge. Meanwhile, the second component, the discriminator, verifies whether the product is genuine or counterfeit. In this way, discriminators continue to challenge generators and create more realistic products. Deepfake technology (GAN) was developed by computer scientist Ian Goodwell with the goal of using machine learning to create the most realistic images possible. But when used inappropriately by the general public or those who don't understand ethics.Ethical concerns for deepfakes include pornography, illegal commercial use, and politics. There are also some questions about the right to speak face-to-face after death. Simply using deepfakes can create unique scenarios that can lead to misinformation that can endanger others.
There are some ways where we can prevent the spread of incorrect information all across the internet. For now, we can detect if a video or image is edited with our bare eyes. But then there are concerns about how Deepfake can produce a such product where it will be perfectly realistic to our eyes. Other than that, companies such as Microsoft and Facebook created a challenge called DFDC that encourages others to participate in the creation of the technology that can be used to detect if an AI has been used to manipulate a video or image that might lead to fake news.
