TL;DR: A new prompting style enables builders to generate clean, poster-like portrait compositions with precise graphic layouts.
Summary: AI creator @xiaoxiaodong01 shared a specific prompting technique for generating editorial-style image-text composite portraits. The method uses high-brightness flat color fields, cropped subjects entering from frame edges, and minimal graphic symbols placed above the subject. The technique is optimized for rendering detailed skin textures and pores while maintaining a minimalist aesthetic suitable for posters.
Why it matters: This provides a reliable template for programmatic layout design in text-to-image pipelines. Developers can adapt these composition rules to automate the creation of high-quality marketing assets and UI graphics.
Source: @xiaoxiaodong01
原文 (Original):
GPT2 x 人像 x 拼贴 x 像素 x 美学提示词 看着十字绣想到的这个创意 其次， 没想到GPT2也知道明星 美在哪里 提示词放在评论区，欢迎交作业返图 | GPT2 x 人像 x 面部三角 x 美学提示词 今日最佳人像提示词， 捏出来了 版面特点就是这个三角，还有黑白的背景逻辑 提示词我放评论区了，使用后，欢迎返图 ⤵️ | GPT2 x 人像 x 窄缝 x 悬疑 x 美学提示词 基于瞳孔