一个有趣的 GPT Image V2 时尚提示词，我将其用于 4 位著名战士。一个有趣的实验。

A fun GPT Image V2 fashion prompt that I used for 4 famous warriors. A fun experiment. 2x2 grid, 16:9, do this for 4 famous warriors: <instruction> Input A is the subject photo. You are a god-tier fashion designer, costume director, textile artist, and editorial photographer. Phase 1: Wardrobe Fusion Randomly select TWO Fashion Archetypes and TWO Textile Systems. Blend them into a coherent wearable identity. Fashion Archetypes: [Avant-Garde Runway Couture, Nomadic Desert Robe, Cybernetic Streetwear, Victorian Mourning Dress, Samurai Armor Silhouette, 1970s Disco Glamour, Arctic Expedition Gear, Renaissance Court Attire, Punk Deconstruction, Monastic Minimalism, Space-Age Mod Fashion, Baroque Opera Costume, Techwear Utility Suit, Ballet Rehearsal Layers, Royal Ceremonial Regalia] Textile Systems: [Quilted Panel Construction, Laser-Cut Mesh, Hand-Embroidered Appliqué, Reflective Foil Laminate, Heavy Wool Draping, Transparent Organza Layers, Ribbed Knit Compression, Modular Buckle Harness, Beaded Fringe Motion, Pleated Architectural Folds, Weatherproof Sealed Seams, Chainmail Textile, Printed Silk Illustration, Inflated Puffer Volumes, Raw Frayed Edges] Phase 2: Visual Rendering Generate a 2:1 horizontal asset sheet. LEFT HALF: Transform the subject’s clothing, accessories, and styling using the generated fashion identity. Preserve exact body pose, face angle, silhouette boundaries, background layout, and camera framing. RIGHT HALF: Create a dark fashion-spec panel with clean white text. Phase 3: Text Content Render this text on the right panel: "Look Name: [Generate a name from the 4 selected elements] <instruction> Input A is the source photo. Goal: Restyle the subject into '[Look Name]'. Rules: Pose Fidelity: Preserve exact body pose, silhouette, camera angle, and facial direction. Fashion Fusion: Combine [Fashion Archetype 1] with [Fashion Archetype 2]. Textile System: Construct garments using [Textile System 1] and [Textile System 2]. Details: Include seams, closures, textile weight, movement, accessories, and styling logic. Lighting: Editorial fashion lighting suited to the materials. Rendering: Photorealistic high-fashion campaign image. Output: ONE image, high-fidelity wardrobe transformation. </instruction>" </instruction>

一个有趣的 GPT Image V2 时尚提示词，我将其用于 4 位著名战士。一个有趣的实验。 2x2 网格，16:9，为 4 位著名战士执行此操作：<instruction> Input A 是主体照片。你是一位神级时装设计师、戏服总监、纺织艺术家和时尚大片摄影师。 第 1 阶段：衣橱融合 随机选择 TWO（两个）时装原型和 TWO（两个）纺织系统。将它们融合成一个连贯的可穿戴身份。 时装原型：[Avant-Garde Runway Couture（先锋秀场高定）, Nomadic Desert Robe（游牧沙漠长袍）, Cybernetic Streetwear（赛博街头服饰）, Victorian Mourning Dress（维多利亚式丧服）, Samurai Armor Silhouette（武士盔甲廓形）, 1970s Disco Glamour（1970 年代迪斯科魅力）, Arctic Expedition Gear（北极探险装备）, Renaissance Court Attire（文艺复兴宫廷服饰）, Punk Deconstruction（朋克解构）, Monastic Minimalism（修道极简主义）, Space-Age Mod Fashion（太空时代摩登时尚）, Baroque Opera Costume（巴洛克歌剧戏服）, Techwear Utility Suit（机能风工装套装）, Ballet Rehearsal Layers（芭蕾排练层搭）, Royal Ceremonial Regalia（皇家仪式礼袍）] 纺织系统：[Quilted Panel Construction（绗缝拼接结构）, Laser-Cut Mesh（激光切割网眼）, Hand-Embroidered Appliqué（手工刺绣贴花）, Reflective Foil Laminate（反光箔层压）, Heavy Wool Draping（厚羊毛垂坠）, Transparent Organza Layers（透明欧根纱层叠）, Ribbed Knit Compression（罗纹针织压缩）, Modular Buckle Harness（模块化扣具背带）, Beaded Fringe Motion（串珠流苏律动）, Pleated Architectural Folds（建筑感褶裥）, Weatherproof Sealed Seams（防水密封接缝）, Chainmail Textile（纺织锁子甲）, Printed Silk Illustration（印花丝巾插画）, Inflated Puffer Volumes（充气蓬松体积）, Raw Frayed Edges（原始毛边）] 第 2 阶段：视觉渲染 生成 2:1 水平资产图。 左半部分：使用生成的时尚身份转换主体的服装、配饰和造型。精确保留身体姿势、面部角度、廓形边界、背景布局和镜头取景。 右半部分：创建一个深色时尚规格面板，文字为干净的白色。 第 3 阶段：文本内容 在右侧面板上呈现以下文字： "Look Name: [根据选定的 4 个元素生成名称] <instruction> Input A 是源照片。目标：将主体重新造型为 '[Look Name]'。规则：姿势保真度：精确保留身体姿势、廓形、镜头角度和面部朝向。时装融合：将 [Fashion Archetype 1] 与 [Fashion Archetype 2] 结合。纺织系统：使用 [Textile System 1] 和 [Textile System 2] 制作服装。细节：包含接缝、闭合方式、面料重量、动态感、配饰和造型逻辑。光线：契合材质的时尚大片编辑级布光。渲染：照片级真实感的高端时尚广告图像。输出：ONE（一）张图像，高保真衣橱改造。 </instruction>" </instruction>