Trending Deepfake Files of Singing Video Portraits Goes Viral

Imperial College in London and Samsung’s AI research center in the UK recently tracked upon new research regarding the emergence of deepfake publications. Discoveries of single photos and audio files generated into singing or talking video portraits are developed and uploaded to the public through youtube and other social networking sites. Similar to other deepfake programs, the researchers utilize machine learning to generate their output. Although the results of these creations are not completely realistic, in a sense, they are highly complimented upon when considering how little data is needed to create one and result with notable outcomes.

Combining various photos and audio files of famous celebrities, ordinary people and music make these deepfakes more relatable to society. Here are some examples that will entertain:

This video is a real clip of Albert Einstein speaking by incorporating a clip from his voice. With a little bit of configuration and imagination, a 16-second lecture is created.

In this amusing clip, Grigori Yefimovich Rasputin is orchestrated of belting out the Beyoncé classic ‘Halo’ surprisingly, perfectly.

Lastly, a more realistic approach to creating a deepfake file is displayed in this very clip of generating a video that not only matches the input audio but is tweaked to communicate a specific emotion that is limited in utilizing with a single photo and an audio file.

What Is Deepfake?

Display Image


Deepfake, also known as ‘deep learning’ or ‘fake’, is a technique for human image synthesis based on artificial intelligence. It is used to combine and superimpose existing images and videos onto source images or videos using a machine learning technique. Because of these capabilities, deepfakes have been inappropriately and negatively manipulated to create fake celebrity pornographic videos or revenge porn. Deepfakes can also be resorted to create fake news and malicious jokes. Often the development of deepfakes is accounted upon two distinct settings: research at academic institutions and development by amateurs in online communities. Academic research related to deepfakes lies predominantly within the field of computer vision, often grounded in artificial intelligence that focuses on computer processing of digital images and videos. In contrast, amateur creators of certain deepfakes produce and share online within their communities (Reddit) of contents involving celebrity face swaps onto pornographic actresses, etc.

These work of deepfake files demonstrate how quickly this technology is progressing. Techniques for generating deepfakes have become proficient and easier to access every day.

Because numerous deepfakes are referenced of bad reputations, it is no wonder the discovery of this new research in the Imperial College and Samsung’s AI research center is highly praised for its effort in appeasing a light relief to society, regardless of how weird and terrible things might get in the future.

Here Is Why Zendaya’s HBO Show Euphoria Is A Career Changer
Here Is Why Zendaya’s HBO Show Euphoria Is A Career Changer
Read More:
  • 10614935101348454