Remember your astonishment when you first saw real-time AR-based filters on Snapchat? This wore out quickly, now it’s a given. Do you recollect these filters that transformed your photos as if they were painted by Van Gogh, Dali, or Klimt? These are also familiar. Now blend the two and get the AI-generated selfies that combine the best of both worlds.
What will an AI selfie generator do? First, it will learn how you look. This will take some of your selfies, the way they are – photos where your face prevails, and the background doesn’t interfere much. This will be enough for the AI to rebuild your appearance in 3D, so it can model it from various POV. Then choose the style and see how it fits your image.
The greatness of this is that your resulting photos will not be bound to your actual posture and environment, face expressions, or mood. This is a refined version of you put into various circumstances and then redrawn. This gives you incredible new opportunities when it comes to anything from creating new avatars for social media to reinventing yourself as a person.
There are more than one tools for doing this already. The most popular, probably because it’s free, is DrawAnyone, which requires five photos of a person (you, your significant other, or whoever’s photos you store on your devices). It will require an hour of waiting before the system processes the pictures, and then you’re good to go. The site can make it an oil portrait, a digital painting, or a portrait of an anime character.
What about the uncanny valley effect some may already anticipate? The answer: it depends. Some pictures are really realistic, but some have that robotic sense of unnaturalness which is yet to be overcome. The greatest thing about AI, though, is that it’s learning, and soon this problem may be gone.
Are you impressed with these AI-powered selfies? Which style did you like the most? Will you share some good sites that make it better? We’d like to read it in the comments!
