OpenZhenZhen AI 提示词画廊 · gpt-image-2.5 · image

原主体与原生艺术风格的上下双拼

Original Subject with Art Brut Style in Vertical Diptych

一张3:4竖幅完整插画，上下各占画布一半：上半清晰、自然地展示用户原图中的主要人物或常见宠物；下半将同一主体转译为粗粝、手作感的Raw Art / Art Brut原生艺术。保持主体身份可辨、构图完整，背景简洁透气，不添加文字或界面元素。

原主体与原生艺术风格的上下双拼 — AI image creation
原主体与原生艺术风格的上下双拼

提示词 / Prompt

中文提示词

以用户上传的参考图作为主要可见人物的面部身份依据。在不与身份参考冲突的情况下，保留来源的场景、造型、服装、姿势、构图、光线和视觉风格。 创作一张完整的3:4竖幅插画，严格在画布正中水平分为上下两半。整体是一件连贯作品，不是把两张独立图片拼贴在一起。使用用户上传的原图作为唯一主要主体的来源。若主体是人物，保留其可辨识特征，并对面部身份使用已配置的ref-10a-v1；不要把人物身份参考套用到动物身上。若主体是猫、狗或其他常见宠物，则保留原图中同一动物可辨识的毛色与花纹、体型和耳形、眼睛及神态。不要添加无关主体。 上半部分——原主体展示：清晰、自然并尽可能写实地呈现同一主体，视觉中心稳定，主体完整入镜，避免意外裁切。人物需保留可辨识特征、年龄呈现、服装和配饰；宠物需保留其独特外观。复杂建筑、室内空间或街景只作为次要背景适度简化，不将其作为重点。上半部分明确承担原图主体展示功能。 下半部分——Raw Art / Art Brut原生艺术转译：将完全相同的人物或动物转化为完整且仍可明确辨认的原生艺术插画。使用未经学院化修饰的线条、粗粝形体、直接的符号标记、不规则色块、富有表现力的墨迹笔触和直接的情绪张力。可在绘画风格中有意识地夸张或扭曲，但不得改变主体身份。构图保持一个明确主体或场景、视觉中心稳定、主体完整不裁切，背景简洁并保留呼吸感。线条、色彩、明暗、材质和可见纸张纹理须统一服务于手作风格；避免混入无关画风、过度光滑、无意义细节堆积和随机装饰。 上下两半必须清楚呈现同一来源主体，形成“原主体 / 风格化主体”的对照。下半部分应质朴、原始、有冲击力并具有手作感，不要变成精致商业插画，同时保留核心辨识度。除非主体表达确有需要，不添加文字、Logo、水印、边框、UI或额外拼贴画格。

English prompt

Use the uploaded reference image as the facial identity reference for the main visible person. Keep the source scene, styling, clothing, pose, composition, lighting and visual style unless they conflict with the identity reference. Create one complete 3:4 vertical illustration with a strict horizontal top/bottom split at exactly half height. It is a single integrated artwork, not two separate images pasted together. Use the uploaded original image as the source for the sole main subject. If that subject is a person, preserve their recognizable traits and use configured identity ref-10a-v1 for facial identity; do not apply the human identity reference to an animal. If the subject is a cat, dog or other common pet, preserve that same animal's identifiable fur color and pattern, body and ear shape, eyes and expression from the original image. Do not add unrelated subjects. TOP HALF — Original-subject display: show the same subject clearly, naturally and as realistically as possible, with a stable visual center and the full subject visible without accidental cropping. Preserve identifying features, age presentation, clothing and accessories for a person, or distinctive appearance for a pet. Simplify complex buildings, interiors or streets into secondary background context rather than making them the focus. Make this upper half the clear source-subject view. BOTTOM HALF — Raw Art / Art Brut translation: transform that exact same person or animal into a complete, unmistakably recognizable Art Brut illustration. Use unrefined, non-academic lines, rough forms, direct symbolic marks, irregular color blocks, expressive ink strokes and emotionally direct visual tension. Allow deliberate exaggeration and distortion in the drawn style without changing which person or animal it is. Keep one clear subject or scene, a stable center, full uncropped subject, and a simplified breathable background. Make lines, colors, light and shadow, materials and visible paper texture belong to this one handmade style; avoid unrelated mixed styles, excessive smoothness, meaningless detail piles and random ornament. Both halves must clearly depict the same source subject as an original-versus-stylized comparison. The lower half should feel handmade, simple, primitive and impactful rather than polished commercial illustration, while retaining core recognizability. Unless essential to the subject, add no text, logos, watermarks, borders, UI or extra collage panels.

查看 API 模型

探索更多图片和视频提示词 · 生成图片 · 生成视频 · 在 AI 画布中创作