This year I was very lazy to design my own Easter Greeting Cards with Illustrator and Photoshop so I've asked Stable Diffusion to do the job for me (*shame on me*). Generative Art is THE big thing in 2023 and everybody is only speaking in terms of "prompts" (Chat-GPT, anyone?). Prompts are small words sequences or even sentences you can give a generative model as an input and then with the magic of AI aka the model behind it an output will generate a picture for your. The visual quality of the output depends on the parameter (here the prompts/words) you give to the model as an input.
Generating art or picture is not something new, but thanks to the recent advances in Deep Learning and the text-to-image models almost everyone can generate his own "AI-Art" within a few clicks. Popular models are Stable Diffusion, Midjourney and its "open-source"-version Open Journey in those you can get some insights when visiting sites like HuggingFace (https://huggingface.co)
Although the generation task can be locally done on a computer with the help of ready-to-use user interface like Automatic1111 (https://github.com/AUTOMATIC1111/stable-diffusion-webui/wiki/Install-and-Run-on-NVidia-GPUs) (though it takes a very long time if you don't have a GPU), for the purpose of this tutorial we are going to use the website Patience.ai and Craiyon, so that you can directly enter your prompts and see the result a few seconds later.
Step 1: Go to Patience.ai
Patience.ai (https://www.patience.ai/models) is a website that enables visitors to test a huge amount of Stable Diffusion and Open Journey models to generate AI art. Select the model "Dreamshaper" (https://civitai.com/models/4384/dreamshaper) which was fine-tuned by the Stable Diffusion AI-Art community and which delivers pretty good results if you want to generate dreamy/cute stuff. Press on "Start" and you will get to a next page that will ask you to enter a prompt.
Step 2: The prompt
Paste following prompt:
very cute and fur Easter rabbit, pixar, disney, cute eyes, happy, smiling, lot of colorful eggs around,
in a basket, vegetal, background, bloom, lense, 4K, fur, photorealistic
And press the "Dream" button.
After a few seconds/minutes you will get your first results: a lot of cute rabbits.
If you are not completely happy with the output you can adjust some settings and create newer variations of your initial picture.
Now you can also fine tune a selected result and add some variation in the parameters of the model. input
Step 3: Admire your (funny/weird) results
I've played a few minutes with Patience.ai and tada here is a small selection of generated images for the given prompt.
.png)
.png)
Final step: Upgrading the generated picture
In this last step, we really need to use Photoshop to add the "Happy Easter Day" text, as generative models have currently huge difficulties to produce coherent text fonts. And here is the result:
Looks cute, isn't it? But have you noticed that the basket has a curious form and texture? But seriously who would notice this? It's Easter and we are focussed on finding chocolate Easter eggs! Happy Easter!
Alternative: Craiyon v2
If you are looking for more wacky and somehow funny results, then try Craiyon! The placement of the rabbit's eyes gives the poor animal a very stupid and naive look!
