{"id":6639,"date":"2024-03-29T11:26:38","date_gmt":"2024-03-29T03:26:38","guid":{"rendered":"https:\/\/www.1ai.net\/?p=6639"},"modified":"2024-03-29T11:26:47","modified_gmt":"2024-03-29T03:26:47","slug":"gpt%e7%bb%98%e5%9b%be%ef%bc%8c%e5%a6%82%e4%bd%95%e7%94%a8chatgpt%e7%bb%98%e5%9b%be%e6%97%b6%e4%bf%9d%e8%af%81%e4%ba%ba%e7%89%a9%e7%9a%84%e4%b8%80%e8%87%b4%e6%80%a7%ef%bc%9f","status":"publish","type":"post","link":"https:\/\/www.1ai.net\/en\/6639.html","title":{"rendered":"GPT drawing, how to ensure the consistency of characters when drawing with ChatGPT?"},"content":{"rendered":"<p class=\"pgc-p\" data-track=\"166\" data-pm-slice=\"1 1 []\">How to use<a href=\"https:\/\/www.1ai.net\/en\/tag\/chatgpt\" title=\"[View articles tagged with [ChatGPT]]\" target=\"_blank\" >ChatGPT<\/a><a href=\"https:\/\/www.1ai.net\/en\/tag\/%e7%bb%98%e5%9b%be\" title=\"_Other Organiser\" target=\"_blank\" >Drawing<\/a>How to ensure the consistency of the characters?<\/p>\n<p class=\"pgc-p\" data-track=\"167\">The output of DALL-E that comes with chatgpt is unstable, including unstable character consistency and unstable aspect ratio. Today, I will teach you the simplest way to achieve stable aspect ratio and character consistency, so that you can easily start hand-drawn book content production, lower the threshold, and quickly get positive and negative feedback.<\/p>\n<p class=\"pgc-h-arrow-right\" spellcheck=\"false\" data-track=\"137\">1. Problem Controlling Aspect Ratio<\/p>\n<p data-track=\"138\">Midjourney's friends know that we can control the size of the picture we want by order of the \u2013ar, and it's not as good as it seems in DALL-E\u3002<\/p>\n<p data-track=\"139\">Currently, DALL-E supports 3 resolutions:<\/p>\n<ul>\n<li data-track=\"140\">Square (1024 x 1024): This is the default resolution, which is automatically exported by the system without special requirements for the hint\u3002<\/li>\n<li data-track=\"141\">Screen (1792 x 1024): Images suitable for viewing, panoramic or any level of need, suitable for the production of cross-screen content\u3002<\/li>\n<li data-track=\"142\">Stationary (1024 x 1792): is best suited for the whole body portrait, high structure or any image requiring a vertical direction and suitable for the production of vertical content\u3002<\/li>\n<li data-track=\"143\">So how do you write prompt words to stably generate the desired image size? Start from scratch~<\/li>\n<\/ul>\n<p data-track=\"145\">First I have no ideas, let gpt generate ideas for me.<\/p>\n<div class=\"pgc-img\"><img loading=\"lazy\" decoding=\"async\" class=\"alignnone size-full wp-image-6641\" title=\"get-64\" src=\"https:\/\/www.1ai.net\/wp-content\/uploads\/2024\/03\/get-64.jpg\" alt=\"get-64\" width=\"838\" height=\"771\" \/><\/div>\n<p data-track=\"146\">Just let him output the picture according to prompt 2~<\/p>\n<div class=\"pgc-img\"><img loading=\"lazy\" decoding=\"async\" class=\"alignnone size-full wp-image-6643\" title=\"get-66\" src=\"https:\/\/www.1ai.net\/wp-content\/uploads\/2024\/03\/get-66.jpg\" alt=\"get-66\" width=\"1024\" height=\"1024\" \/><\/div>\n<p data-track=\"147\">as you can see, the direct result is 1024 x 1024 square images. how does that make him a screen? add a keyword: full body portrait or vertical images<\/p>\n<div class=\"pgc-img\"><img loading=\"lazy\" decoding=\"async\" class=\"alignnone size-full wp-image-6642\" title=\"get-65\" src=\"https:\/\/www.1ai.net\/wp-content\/uploads\/2024\/03\/get-65.jpg\" alt=\"get-65\" width=\"780\" height=\"754\" \/><\/div>\n<p data-track=\"148\">as you can see, 1024 x 1792 stand-up pictures have been steadily generated, how do you generate a screen map? use keyword: wide-banded images<\/p>\n<div class=\"pgc-img\"><img loading=\"lazy\" decoding=\"async\" class=\"alignnone size-full wp-image-6644\" title=\"get-67\" src=\"https:\/\/www.1ai.net\/wp-content\/uploads\/2024\/03\/get-67.jpg\" alt=\"get-67\" width=\"803\" height=\"594\" \/><\/div>\n<p data-track=\"149\">At this point, the problem of image size stability is solved.<\/p>\n<p class=\"pgc-h-arrow-right\" spellcheck=\"false\" data-track=\"150\">2. How to solve the problem of character consistency?<\/p>\n<p data-track=\"152\">Method 1: The style of images generated in the same latent space can remain consistent.<\/p>\n<p data-track=\"153\">In layman&#039;s terms, it means to let dall-e generate a multi-grid image. For example:<\/p>\n<div class=\"pgc-img\"><img loading=\"lazy\" decoding=\"async\" class=\"alignnone size-full wp-image-6645\" title=\"get-68\" src=\"https:\/\/www.1ai.net\/wp-content\/uploads\/2024\/03\/get-68.jpg\" alt=\"get-68\" width=\"795\" height=\"559\" \/><\/div>\n<div class=\"pgc-img\"><img loading=\"lazy\" decoding=\"async\" class=\"alignnone size-full wp-image-6646\" title=\"get-69\" src=\"https:\/\/www.1ai.net\/wp-content\/uploads\/2024\/03\/get-69.jpg\" alt=\"get-69\" width=\"770\" height=\"537\" \/><\/div>\n<div class=\"pgc-img\"><img loading=\"lazy\" decoding=\"async\" class=\"alignnone size-full wp-image-6648\" title=\"get-71\" src=\"https:\/\/www.1ai.net\/wp-content\/uploads\/2024\/03\/get-71.jpg\" alt=\"get-71\" width=\"782\" height=\"565\" \/><\/div>\n<p data-track=\"154\">After that, crop and enlarge the high-definition image, and you can start creating.<\/p>\n<p data-track=\"155\"><strong>Method 2:<\/strong><\/p>\n<p data-track=\"156\">If you want to control the performance of each graph, you can use the following method:<\/p>\n<p data-track=\"157\">Use prompt words: upper left, lower left, upper right, lower right layout segmentation<\/p>\n<p data-track=\"158\">Please note that this is one image, not the four images that DALL-E 3 generates by default.<\/p>\n<p data-track=\"159\">Prompt word template: [Medium] [Layout] [Upper left description] [Upper right description] [Bottom left description] [Bottom right description]<\/p>\n<div class=\"pgc-img\"><img loading=\"lazy\" decoding=\"async\" class=\"alignnone size-full wp-image-6647\" title=\"get-70\" src=\"https:\/\/www.1ai.net\/wp-content\/uploads\/2024\/03\/get-70.jpg\" alt=\"get-70\" width=\"793\" height=\"601\" \/><\/div>\n<p data-track=\"160\">Finally, by analogy, can you let dall-e generate a story in one go?<\/p>\n<div class=\"pgc-img\"><img loading=\"lazy\" decoding=\"async\" class=\"alignnone size-full wp-image-6649\" title=\"get-72\" src=\"https:\/\/www.1ai.net\/wp-content\/uploads\/2024\/03\/get-72.jpg\" alt=\"get-72\" width=\"1024\" height=\"1024\" \/><\/div>\n<p data-track=\"161\">The layout of the picture determines the size, multiple grids plus the description of the layout, can you still say that the consistency is not good? Or do you think it is difficult to split and enlarge it?<\/p>\n<p data-track=\"162\">Of course, the gameplay shown in the picture above can be extended to many other ways. For example, is it possible to make several pictures into frame pictures of a person dancing, and then edit them after processing?<\/p>\n<p data-track=\"163\">For example, is it possible to generate a frame-by-frame image of a person or a celebrity&#039;s facial expression, from calm to big to crying, and so on?<\/p>\n<p data-track=\"164\">For example, is the entire process of a dragon opening its mouth and breathing fire possible?<\/p>\n<p data-track=\"165\">Of course, the gameplay of dall-e is far more than that. Go ahead and explore it, young man. AI is a tool, and the tool allows you to fiddle with it at will. As for how to apply it to money-making scenarios, this is the key.<\/p>","protected":false},"excerpt":{"rendered":"<p>How can you use ChatGPT to map people for consistency? Using chatgpt's own DALL-E displays of instability, which include the unsatisfied, unsatisfied, unsatisfied and unsatisfied, teaching people today the simplest way to achieve broad-symmetric stability and consistency in character, so that you can easily draw the content, lower the threshold and get the positive feedback. First, it's a matter of mind control, using Midjourney's friends, to know that we can control the size of the picture we want through the \u2013ar command, and it's not as good as it seems in DALL-E. For the time being, DAL-E supports 3 resolutions: square (1024 x 1024): this is silent<\/p>","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[149,144],"tags":[177,1964],"collection":[258],"class_list":["post-6639","post","type-post","status-publish","format-standard","hentry","category-jiaocheng","category-baike","tag-chatgpt","tag-1964","collection-chatgpt-prompt-guide"],"acf":[],"_links":{"self":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/6639","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/comments?post=6639"}],"version-history":[{"count":0,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/6639\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/media?parent=6639"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/categories?post=6639"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/tags?post=6639"},{"taxonomy":"collection","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/collection?post=6639"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}