Researchers from Cambridge Consultants has trained an AI to interpret simple doodles and turn them into artistic masterpieces.
The trained machine is named Vincent. Learning from less than 10,000 images using a global adversarial network, Vincent is an AI that works by detecting counterfeits, and from that it can produce something similar to arts from the Renaissance period.
Vincent isn't created to produce art. Instead, it imitate the styles of doodles and tries to complete the image in a form of a painting.
According to Monty Barlow, director of machine-learning at Cambridge Consultants:
Related: Google’s AI Is Able To Turn Doodles Into Proper Pictures
In short, Vincent is just an AI that doesn't express emotions or feelings. The researchers themselves aren't interested in creating Vincent as a better artist, or becoming a better art creator.
Here, machine-learning developers are trying to create a more rational, logical, and sane robot capable of figuring things out on its own.
So rather than being told to do, researchers are trying to make computers to learn on their own, and in Vincent's case, using an adversarial network to train the AI by giving it conflicting goals
So Barlow and his team isn’t concerning art. Instead, they want to create AI that can better see and understand its surroundings. This is crucial to nearly every aspect of autonomy.
Related: AI Tool Turns Simple Drawing Doodles Into Horror Portraits