Artificial intelligence in painting

Main article: Artificial Intelligence

2024: A robot artist capable of copying the canvases of famous masters was presented in Russia

Russian scientists from the St. Petersburg State Electrotechnical University "LETI" have developed a robot artist capable of reproducing the technique of applying strokes of famous painters when creating copies of paintings. This became known on October 1, 2024. Read more here.

2022: Neural network learned to draw with the artist

Cloud platform Yandex Cloud, studio Arka and multimedia artist Andrew Berger presented a neural network co-author. Yandex announced this on December 8, 2022. Read more here.

2021: Nvidia unveils GauGAN AI system to convert text to image

On November 24, 2021, it became known that the Company Nvidia introduced a technology-based system (artificial intelligenceGauGAN2 successor to the first GauGAN model), which allows you to create believable photographs of non-existent landscapes. With techniques such as segmented mapping, retouching, and text-to-image conversion, GauGAN2 is able to create realistic images based on text and handwritten sketches. More. here

2019: Microsoft unveils art-based image generator

In early March 2019, Microsoft announced a project to generate images based on works of art. To create this service, the developers used the microservice architecture of deep neural networks, Azure services and BLOB object storage. Visual Studio Code and Azure Kubernetes Service allow you to create new images in real time and are responsible for the interactive appearance of the site. Read more here.

2018: AI portrait sold for $433K

In October 2018, a painting painted by artificial intelligence was sold at Christie's auction in New York. For "Portrait of Edmond Belami" from the "Belami Family" series, which was created on the basis of paintings by famous painters, they paid $432.5 thousand at the original lot price of $7-10 thousand. Read more here.

2016: Google representatives helped out from psychedelic paintings written by AI, about $98 thousand

In 2015, a Google team tested neural networks for the ability to independently create images. Then artificial intelligence was trained on the example of a large number of different pictures. However, when the car was "asked" to depict something on its own, it turned out that it interprets the world around us somewhat strangely. For example, on the task of drawing dumbbells, the developers received an image in which the metal was connected by human hands. This probably happened due to the fact that at the training stage the analyzed pictures with dumbbells contained hands, and the neural network misinterpreted this.

On February 26, 2016, in San Francisco, at a special auction, Google representatives rescued about $98 thousand from psychedelic paintings painted by artificial intelligence. These funds were donated to charity. One of the most successful paintings of the machine is presented below.

A picture painted by Google artificial intelligence. Source: factroom.ru

In early 2016, with the help of such technology, the painting "The Next Rembrandt" was painted. Researchers of the Next Rembrandt project in the course of their work analyzed about 350 paintings by the great artist, using 3D scanners that allowed the neural network to capture even the smallest details of the work and copy the writing style of all works.

At the second stage, scientists determined that specifically the neural network should write on its own. For this, they chose a man of 30-40 years old, who has hair on his face, collar and hat.

Portrait "The Next Rembrandt"

The results leave a twofold feeling. On the one hand, we all see that we have a car in front of us. On the other hand, if we put several paintings of Rembrandt nearby, then few people will be able to distinguish this particular portrait.

Microsoft and MIT have created an AI algorithm to find relationships between art objects

On August 18, 2020, it became known that researchers from Microsoft and the Massachusetts Institute of Technology (MIT) developed the MosAIc system to find relationships between art objects from various cultures and eras. Read more here.