Using the artificial intelligence of Stable Diffusion, an artist has developed a code allowing to extract color palettes from a text.
To achieve this result, the London artist Matt DesLauriers showed some elbow grease by combining artificial intelligence and JavaScript.
Stable script delivery
While some artists ban works created by AIs like Mid-journey or Stable Diffusion, others embrace what these tools can bring to art.
This is particularly the case of Matt DesLauriers, who took advantage of the capabilities of Stable Diffusion, an open source tool released less than a month ago and capable of creating synthetic images from a text written by the user, drawing inspiration from millions of images gleaned from the Internet.
Pushing the idea further, the London artist has developed a JavaScript code called “gifenc” to analyze the text behind the image generated by the AI and to extract a color palette. To achieve this result, Matt indicates that it is necessary to have a local installation of Stable Diffusion, as well as a Node.JS.
An embryonic but already effective project
Still in the prototype state, the code created by the artist has been shared on a dedicated GitHub page, cited in source below. On his Twitter account, Matt DesLauriers demonstrates its effectiveness through various tests.
By slightly altering the original text, the code is able to display slightly different colors, aiming to reflect the author’s intent. If the idea seems trivial at first glance, the ability to extract a color palette from a text can have very concrete applications.
In particular, this can help artists choose colors that go well together to present a setting or convey a concept. For example, it would be possible to represent in color “the gloom of Monday morning” or “a crossover between Star Wars and The Lord of the Rings “.
It should be noted, however, that Matt DesLauriers’ project is not the first in its category. Last May, an artist nicknamed “dribnet” made available to everyone an online tool capable of producing a palette of six colors from a text.
We bet that other artists versed in code and artificial intelligence will add their stone to the building to develop this original way of conceiving art through the prism of technology.
Source : GitHub
11