How to use AI to Combine Your Personal Style with Architecture and Interior Design
Introduction
We are now living in a world where design concepts from completely different fields can be combined with a few keystrokes to create one-of-a-kind design workflows. With AI tools like Midjourney at our fingertips, we can imagine all sorts of crossover ideas. From a Lamborghini-designed line of kitchen gadgets, to beds from unexpected designers, to shoes that will hopefully never exist, we’re at the verge of a revolution in cross-pollination between disciplines. So as an example, how about we explore the intersection of your personal style with architecture and interior design through a mainstay of fashion theory: seasonal color analysis. You may know whether you’re a winter, summer, autumn, or spring, but what season is your dream home?
A prevalent theory in the world of personal style is a 12 type color palette system used for seasonal color analysis. It is organized by the four seasons, which are further subdivided into 3 subtypes based on the “chroma” (intensity/saturation) and “value” (lightness/darkness) of their colors (See citations [1] and [2]).
Architects tend to analyze color with a very different, less well defined theoretical framework for their color theory (See citations [3], [4], [5], and [6]) so ideas like seasonal analysis are not often applied to architecture. However if you have a personal style and palette you like, with this quick AI workflow you can discover new ways of translating that style to something you can express in through your design choices for your home.
AI Experiments in Seasonal Color Analysis
To study these possibilities, I turned to ChatGPT for some word association help, asking for an interior design style, room type, and architecture style for each of the 12 season color palettes. To further reinforce the style, I also used a set of nature photo Pinterest boards curated by color theory blog ElementalColor as image prompts to liven up the resulting images with some biophilic motifs more than simply text prompting would. Finally, using Midjourney (an AI image generator perfectly suited for integrating seemingly unrelated concepts like this), I combined Elemental Colour’s pinterest boards with the suggestions from ChatGPT using a prompt template I developed for this experiment. Let’s look at two examples of applying color theory to design – one in interior design, and one in façade design.
To start off, here are our goals courtesy of ChatGPT:
Bright Spring: True Spring: Light Spring: |
Light Summer: True Summer: Soft Summer: |
Dark Winter: True Winter: Bright Winter: |
Soft Autumn: True Autumn: Dark Autumn: |
And here are the seasonal color analysis images from ElementalColour we will be using for inspiration ([ii]), to set the palette and the vibe.
Midjourney Results – Interior Design
Using image prompting and a consistent prompt structure (with a little prompt engineering where needed) let’s see what Midjourney thinks the perfect interior design would be for each seasonal palette.
Which one speaks to you? Leave a comment below with your favorite. I will be doing a future post expanding upon the style that gets the most votes in the comments.
Prompt Format: {IMAGE} Award winning photo of a unique stylish [ROOM], [STYLE] interior design, [SEASON] color palette –no webpage, website, collage, grid, paintings, wall art, photo gallery, [OTHER UNDESIRED RESULTS] –ar 2:1 –s 300 –iw 0.5 –c 20 –style raw –v [VERSION (5.2 OR 6.0)] –seed [SEED]
For those of you who want to make images like this yourself, let’s break down this prompt:
- The image prompt is a link to the Elemental Colour Pinterest photos shown above. It sets the color palette and cues Midjourney to draw inspiration from the photos they contain. For example this results in balmy seaside imagery for Light Summer but faded hazy floral imagery for soft summer. Specifying the season in words also helps reinforce that color palette.
- The negative prompt (–no) removes common issues with this method, such as Midjourney translating the Pinterest grid into a website screenshot or grid of wall art within the scene rather than properly integrating it into the design. Depending on the context, you may need to add more negative prompts if Midjourney latches onto an idea you didn’t want.
- Aspect ratio (–ar) sets a 2:1 landscape format.
- Stylize (–s) sets how much Midjourney should prioritize aesthetics over strict adherence to the prompt.
- Image weight of 0.5 (–iw) sets the importance of the Pinterest image at half the importance of the text prompt (to avoid just generating another Pinterest screenshot).
- Chaos (–c) increases the variability in image composition Midjourney will produce on a scale from 0 to 100.
- “Style raw” instructs Midjourney to use its more literal photographic mode rather than its more abstract painterly default mode.
- Version (–v) sets the version of Midjourney to be used. Each version has its strengths and weaknesses. The images in this study use mainly version 5.2 and its more rendered aesthetic while some use version 6.0 and its more photoreal aesthetic for the styles 5.2 isn’t as good at. Can you spot which ones are version 6.0?
- Seed sets a consistent starting point for all image generations to limit random variation and so the impact of small changes to the prompt can be more easily compared. If a seed is not specified the same prompt can produce completely different results on each try. Seed can be a randomly selected number you use consistently throughout all prompts or the seed of a previous generation where you like the general composition of the result and want to reimagine it with variations.
As always, it is up to the designer to adjust these parameters and engineer the prompt (See Post 4: Prompt Engineering) to better suit the goals of the project.
Midjourney Results – Façade Design
Now that we’ve seen what Midjourney can do for interior design and understand the prompt structure, let’s focus on applying this color theory to the world of architecture and façade design.
Which of these would you want to live in? Just like before, leave a comment below with your favorite to see more of it in a future post.
Prompt Format: {IMAGE} Award winning photo of a unique stylish facade, [STYLE] style architecture, [SEASON] color palette –no webpage, website, collage, grid, paintings, wall art, photo gallery, [OTHER UNDESIRED RESULTS] –ar 2:1 –s 300 –iw 0.5 –c 20 –style raw –v [VERSION (5.2 OR 6.0)] –seed [SEED]
Conclusion
If you would like to see more ideas inspired by seasonal color analysis or any of these styles, leave a comment. If you want to learn more fast and easy AI design workflows like this, you can use the sign-up below to subscribe to the free Pixels to Plans Newsletter,
I hope this shows the possibilities AI opens up in combining different design concepts not usually considered in architecture and demonstrates that design inspiration can come from anywhere. If you can think of any other surprising crossovers between architecture, interior design, and unrelated disciplines, leave a comment at the end of the page.
Also stay tuned for new generative AI design ideas in future posts.
About the Author
AcanthusAlchemist
Designer and engineer exploring the intersection of AI, architecture, and urbanism.
email: acanthus@pixelstoplans.com
Subscribe to our Newsletter
Interested in exploring other AI generated architecture concepts and design workflows? Sign up for our free mailing list below to be notified whenever we post new content here on Pixels to Plans.
Citations:
[1] ElementalColour – The Making of the 12 Tones
[2] The Concept Wardrobe – Color Analysis: A Comprehensive Guide
[3] YouTube – 30×40 Design Workshop – How I use Color (An Architect’s Guide)
[5] YouTube – Upstairs – Colors in Architectural Representation
[6] Foyr – Color Theory Basics: How To Use Color Theory in Interior Design?
Image References:
[i] ElementalColour.com.au – The Making of the 12 Tones
[ii] (Source for all image prompts) ElementalColour.com.au – Images in 12 Tones
I would like a home built in the theme of dark autumn please. Great read!