保持图1中的女性样魏不变，参考图2的化妆细节和发型以及背景，生成一张新的图片1女性的写真。
Want a different result? Change just a few parts
AI finds the swappable parts (subject/style/scene) — fill in what you want, keep the rest
