|
Listen to this story
|
Text-to-image AI art generators have awed the internet for quite some time now. Twitter and Instagram feeds are flooded with art produced with the help of these tools. However, the reigning players in the domain right now are ‘DALL-E 2’ and ‘Midjourney’.
These powerful AI tools create art by transforming words into images.
It is therefore essential to understand how it interprets instructions provided in English. It does not read or comprehend text in the same way as humans do.
Here’s a list of useful prompt generators and guides:
Prompter for MidJourney
Midjourney is one of the latest, free prompt builders for desktop users.
Based on Google Sheets, Midjourney allows Google users to make a copy to save the tool to their drive. Further, the save button copies the current prompt to the user’s ‘My prompts’ sheet along with the date and time.
Note: This action runs via an external script, so when one opens the sheet, it will ask if you want to allow permission to the script. Once accepted, it’ll work every time.
Learn more here.
Disco Diffusion Modifiers
This collection is shared by AI artists KyrickYoung and sureailabs.
Available on Weird Wonderful Art AI, the guide offers a collection of modifiers and styles that can be used for AI art creation.
It consists of modifiers, like artist names, that are keywords which guide the image generation in a certain direction.
Learn more here.
promptoMANIA
Coded and designed by Peter W. Szabo—this is a free prompt builder created to help users make AI art and works in collaboration with Midjourney.
Learn more here.
Artist studies by Remi Durant
Through time, artists have produced art in various media and styles, incorporating different philosophies and ideals in their works. Consequently, different artistic tendencies or styles can be grouped under collective categories of ‘art movements’.
Artist Studies provides a vast spectrum of terms belonging to a range of art movements—from classicism to futurism and abstract expressionism to avant-garde.
A trick to generate impressive art through AI is to include the names of prominent real-world artists. For instance, Remi Durant curated the Artist Studies’ page to explore different styles, colours and subjects that well-known artists approached in their works.
Durant curated the page to help people find AI artist names to add to their prompts.
Artist Studies has nearly 600 artists across four different CLIP models.
Learn more here.
CLIP Prompt engineering for generative art
CLIP was trained on image-text pairs from all over the internet. It is therefore able to create styles based on stereotypical content of online websites or publications.
This blog post by Matthew McAteer provides high-level tips for prompt engineering plus details about CLIP fine-tuned on datasets beyond ImageNet.
Note: If you want to continue your work on a particular prompt output, make sure to save the seed value.
Learn more here.
Disco Diffusion model comparison study
The study compares the results of different combinations of the models available in Disco Diffusion V5.
The first table compares all primary model combinations and all secondary model combinations. The second one performs the same comparison with transposed axes. The rest of the charts are a comparison of each primary and secondary model in isolation.
All of these models can be found and selected under the ‘Models Setting’’ section of the Disco Diffusion V.5 notebook.
Learn more here.
DALL-E prompt book by Dall·ery Gallery
Created by Guy Parsons and published by the AI art website—the DALL·Ery GALL·Ery book is a visual resource designed to inspire creations. It offers valuable pointers on how to get the most out of the AI art generator.
The 82-slide guide covers techniques for demonstrating the results. It covers aesthetics and vibes, recommended adjectives to use and tips on everything from photography to creating painterly portraits and landscapes, film and illustration styles, historical art styles and 3D art.
For photography, it suggests including proximity, angles, lighting quality and even lens type in the prompt.
For illustration, it provides examples of different styles, media and textures.
For more information, click here.
