Intro
With this article I would like to discuss the making of visual artwork with the use of computers assisting the artist by using modern techniques like artificial intelligence and machine learning. How did we get here, what can be done today and where could we go from here.
Using a computer while creating visual art you will come across a couple of different names, styles and software.
Using the terms artificial intelligence or machine learning nowadays is used often to describe the process of teaching algorithms by feeding it huge datasets of images. These algorithms can be used for creating or changing an image you feed the algorithm.
You could be able to let the computer algorithms change a portrait photo into a digital painting resembling the Girl with the pearl earring painted by Johannes Vermeer or the other way around and create a photograph of the girl with the pearl earring.
Other algorithms can turn a photo into a painting, change the style of the image or combine images. Even upscaling a small image with low resolution into a high resolution file is possible using artificial intelligence.
These are all wonderful possibilities for artists to use as their tools and create their work. And to think we are only seeing the first baby steps of these algorithms.
First let’s start with a bit of history.
History
Back in the last century in the 50’s the first computer generated artworks started to appear. These images where mainly based on mathematical calculations to create patterns. With the possibilities of printing and plotting these artworks saw their way onto paper and the walls.
With the arrival of 3D rendering software it became possible for artists to start creating their own realism or surrealism’s. Coming from the movie industry the special effects software also created possibilities creating more and more advanced pieces of work.
Photoscanners and later digital camera’s brought photography to the computer and software like Adobe’s Photoshop gave artists endless possibilities creating and altering photo’s into pieces of art in all kinds of styles.
Software companies always looking for ways to sell more copies of their software kept bringing new algorithms to their editing software and soon enough a digital artist could change the sky in a photograph with the click of a button.
Google and other companies started to create huge datasets of images to teach their systems all about images. Webmasters wanting to check if a real person is filling out a form added captchas to their websites giving the companies behind the captcha’s the ability to let their algorithms become even smarter because of the human curation coming in from the captcha’s functionality on many websites.
Today’s computers uses to create art
Computers can now create realistic looking photographs of people who don’t exist or with deeplearning algorithms exchanging faces between different people. Although these algorithms can be misused and we can already see that starting to happen. For us artists, it gives us new opportunities for creating art and I believe this brings a lot of exiting new art.
With the arrival of Non Fungible Tokens (NFT’s) the digital files can be made unique items, not just a simple copy. This means digital art can be sold where the buyer buys the ownership of a digital artwork. Making use of blockchain technology which is basicly using cryptocurrency Etherium there now is a market for trading these NFT’s and some people think this will be the future of digital art collecting.
Computer assisted or generated art styles
Mathematical art is an art-form where the computer draws using mathematical calculations which can lead to abstract forms like fractals.
Rendering
3D rendering isn’t actually artificial intelligence but it does use physics models to light a scene, the textures in the scene etc.
Using rendering engines it had become possible to create scene’s, places, worlds and even galaxies that do not exist in real life but do resemble what we would expect it to look like because of rendering lights, shadows on textures in a scene.
This next image is full rendered using 3D rendering. If you have read this whole article you already passed a couple of renders.
Rendering a 3d scene was always still done by a human artists but lately the game rendering engines take a lot of work out of the human artists hands by placing objects by using artificial intelligence. Microsoft Flightsimulator 2020 is a nice example of this. This game features the complete world that is populated by trees, buildings, vehicles. The software makes use of geo-data to create a very realistic view of the world.
Lately game engines have become very powerful in creating whole worlds and you are seeing artists excel in using the latest game engines. I suspect this will be a major direction in which the art rendering field will move and it will get momentum now that the metaverse is something that is being build.
Using AI to change art styles and media is also something that is being used by artists creating art. This might save an artist time or even skills because of the computer doing the crunching. Examples of this are photo to painting conversions. Very realistic looking paintings are created with the click of a single button.
AI as a tool for artists
Image editing tools like Photoshop nowadays also had artificial intelligence built in. Replacing the sky in a photograph or click to select a subject. Photoshop is smart enough to guess the subject you want to select most of the times. No need to spend countless hours selecting and masking subjects where a single click nowadays does a similar job.
Topaz has a suite of image editing products that make use of machine learning and artificial intelligence. Enhancing low resolution photo’s into high resolution photo’s is one of its uses. This isn’t just enlarging by adding average pixels, it’s actually trying to see what the subjects are, how contrast and details are used locally and filling in the missing pixels with those best guesses resulting in a high detailed image instead of a blurry simply enlarged image. I use Topaz Gigapixel for this very often and I think it performs way better then Photoshop’s built in rescaler.
Creating computer assisted art using software tools to convert a photo into a painting is also something I use myself often. Most of the time I combine different tools with my own painting and sketching to create the final artwork, but even straight out of these tools an imported photo will look stunning.
Topaz Studio 2 is one of these tools I think is very underestimated and can be used to create all kinds of different kinds of art and styles. You have to really search for this product since it isn’t featured on their website as huge as their AI powered tools, but it performs really well and if you are into using photographs as a base for artworks this is the tool to use. If you want to purchase Topaz Studio 2 or Topaz Gigapixel Ai use the discount code loyal15 to receive a 15% discount on your purchase.
An even easier tool to create a painting from a photo is this amazing action. After starting the photoshop action it will take a while to complete depending on your base image filesize so feel free to grab a cup of coffee. When the action is done your photo will have transformed into an amazing looking artwork with countless layers you can change to create amazing art.
The future of computer generated and assisted art is looking amazing and I can’t wait to see what software engineers will come up with. I think combining these tools where the artist is sort of the conductor of the orchestra of AI tools will deliver the best results creating new styles of art we have never seen before.
Looking for more reading about AI in art creating, this article might be a nice read for you.