作者:Lars Nielsen
翻译:陈超
校对:赵茹萱
本文约1500字,建议阅读5分钟
本文共列举了10条如何为Midjourney编写图像生成提示的建议。
A detailed "cheat sheet" with some keywords to use for better hint image output.
The picture is generated by Midjourney, and the text prompt: beautiful, dreamy city Unreal Engine
For those who have not heard of Midjourney, here is a sentence introduction: Midjourney is a text-image generation APP, similar to OpenAI's DALLE-2 and Stable Diffusion's DreamStudio, using a large number of network images (about 650 million ) - generate stunning images based on the text clues provided. Now in beta - but more features are being added every week...and the future is yet to be imagined! (Currently their AI engine has reached version 3.0).
original prompt text
Anyone who has used Midjourney knows that the more adjectives you have in your prompt, the more vivid and unique your output will be. A typical user prompt, "a cowboy in a tuxedo on the moon", would generate the following pattern:
Image generated by Midjourney - hint: a cowboy in a tuxedo on the moon
But there are more options for creating predictable, consistent image output than naive raw text prompts!
Now directly open the advanced options of text hints!
Here are some ways you can generate image variants, and some advanced settings: (detailed settings follow the table below)
1. Provide keywords - 'style'
2. Stylize
3. Chaos
4. Resolution
5. Aspect ratio
6. Pass the image as a hint as a URL
7. Weighting image prompts
8. Weighting word hints
9. Filter words
In short, your style is affected by the keywords you choose - and you can specify which style you want to make - the broad style classification is as follows -
Style classification in Midjourney - picture from the author
1. Style - Key words
Providing a set of suggested keywords related to a "style" can result in a range of different outputs based on your preferred style. Here are some keywords and art forms/designs/artists/genres based on the style you might want to choose.
Use design/genre as style
Image generated by Midjourney with hints from author
use artist name as style
You can also use Artist as your output style. Here is a sample generated by the same prompt
/ imagine drawing a horse running at sunset in the style of Andy Warhol
Image generated by Midjourney, with artist as style, hints from author
Use rendering/lighting properties as styles
The picture is generated by Midjourney, and the prompt experiment is from the author (prompt text: "Dream Castle")
2. Stylized output
You can add setting --s <some number> (indicates style)
The images below are from the same prompt - plus high and low stylization options.
/imagine firefighters --s 6000
Image generated by Midjourney, tip: firefighter
3. Chaos - increasing abstraction
Increase or decrease the level of abstraction of the topic on a scale of 0-100
/imagine Eiffel tower --chaos 60
Image generated by Midjourney, tip: Eiffel Tower
4. Resolution
To specify the output resolution, you can choose general keywords such as 8K, 4K, realistic, ultra-realistic, ultra-detailed, intricate detail, etc.
Or you can normalize for predictable output hd and quality/--q are two such settings
/imagine red rose flower --hd
/imagine red rose flower --quality 5
5. Aspect ratio
Pinch to explicitly output the width:length ratio of the picture. This is called aspect ratio. The default output is square (1:1 aspect ratio). But if you want a cinematic perspective — or if you just want to be your computer wallpaper, you can adjust the aspect ratio.
/imagine jasmine in the wild flower --ar 4:3
If you want to specify a custom image size, use the following example:
/imagine jasmine in the wild flower --w 600 --h 300
You can't customize the aspect ratio - but you can specify some standard ratios and some non-standard ratios! (here are some examples)
The picture was generated by Midjourney, the text prompt: "Wild Jasmine", the text comes from the author
6. Images as prompts
If you want to get some image-like output, (when you want consistent output) pass in the image URL.
/imagine http://www.imgur.com/Im3424.jpg box full of chocolates
The generated image will take its cues from the seed image (the URL you fill in) and the text hint.
You can upload multiple images as hints
You can specify the weight of the picture (see the next article)
7. Give weight to image hints
If you want your output to look more like your hint image (see above), give this image a higher weight:
(keyword : --iw <number>)
/imagine http://www.imgur.com/ks34f24.jpg chocolates --iw:4
8. Give weight to text hints
/imagine wild animals tiger::2 zebra::4 lions::1.5
9. Filter text from images
Use the --no keyword to remove unwanted subjects from the image:
/imagine KFC fried chicken --no sauce
10. Finally, you can try some interesting keywords
* Sony Alpha α7, ISO1900, Leica M, = to specify any lens type or camera type
* photorealistic , ultra photoreal , ultra detailed, intricate details, = to specify some possible details and a realistic look and less of artistic
* unreal = to specify an unreal engine feel
* volumetric light , cinematic lighting = to specify some lighting conditions
Summarize
We discussed how you can exercise some degree of control over the output images in your Midjourney experiments. But it's not done yet. There is still plenty of room for imagination and experimentation. Experiment with different hints and have fun with your creations, good luck!
Related references:
Creativity in Text-to-Image Generation JONAS OPPENLAENDER, University of Väskylä, Finland https://arxiv.org/pdf/2206.02904.pdf
To try out some text prompts, here is an article (https://pythoslabs.medium.com/lessons-learn-from-using-midjourney-v4-and-some-awesome-images-generated-text-prompts-included-2153dc02db4a ), which contains some remarkable images generated by Midjourney - some example images.
Original title:
An advanced guide to writing prompts for Midjourney ( text-to-image)
Original link:
https://medium.com/mlearning-ai/an-advanced-guide-to-writing-prompts-for-midjourney-text-to-image-aa12a1e33b6
Editor: Wang Jing
Translator profile
Chen Chao , Master of Applied Psychology, Peking University, data analysis enthusiast. I once studied computer science as an undergraduate, and then pursued relentlessly on the path of psychology. In the process of learning, I have increasingly discovered that data analysis has a wide range of applications. I hope to output some meaningful work through what I have learned. I am very happy to join the big family of Data School, keep humble, and keep eager.
Translation Team Recruitment Information
Job content: It needs a meticulous heart to translate the selected foreign language articles into fluent Chinese. If you are an international student of data science/statistics/computer, or are engaged in related work overseas, or friends who are confident in your foreign language proficiency, welcome to join the translation team.
You can get: regular translation training to improve the translation level of volunteers, improve the awareness of the frontier of data science, overseas friends can keep in touch with the development of domestic technology applications, and the background of THU's data-based industry-university-research research brings great benefits to volunteers Development Opportunities.
Other benefits: Data science workers from famous companies, students from Peking University, Tsinghua University and overseas famous schools will all become your partners in the translation team.
Click "Read the original text" at the end of the article to join the Datapai team~
Reprint Notice
If you need to reprint, please indicate the author and source in a prominent position at the beginning of the article (from: Datapi ID: DatapiTHU), and place an eye-catching QR code at the end of the article. If you have an original logo article, please send [article name - official account name and ID to be authorized] to the contact email, apply for whitelist authorization and edit as required.
After publishing, please send the link back to the contact email (see below). Unauthorized reprinting and adaptation, we will pursue their legal responsibilities according to law.
Click "Read the original text" to embrace the organization