带有不同面部表情的肖像网格
这是一个为 GPT-Image 1.5 设计的提示词，用于根据上传的照片创建 3x3 的肖像网格。它要求在生成九种不同面部表情的同时，保持主体形象的精确相似度，并以简洁、现代的美学、柔和的灯光和柔和的背景呈现。
gpt-image-1.5 · 个人资料 / 头像 · 人像 / 自拍 · 摄影 · 社交媒体帖子
中文提示词
以上传的照片为基础参考，创建一个干净、现代的肖像网格，其中包含同一个人不同面部表情和情绪的肖像。将图像以整齐的 3×3 布局排列，并带有圆角和柔和的蜡笔色背景。每张肖像都应保留相同的面部特征、发型和真实感，只改变表情和细微的风格。表情包括：自然微笑、开心的笑、闭眼平静、中性严肃的凝视、手托下巴沉思的姿势、自信沉着的表情、悲伤或失望的情绪、顽皮的撅嘴，以及睁大眼睛的惊讶反应。使用柔和的影棚灯光、高度细致的皮肤纹理、浅景深、服装变化最小化、一致的调色板、超现实的编辑肖像风格、高分辨率和简洁的美学。
原始提示词
Starting with the uploaded photo as a base reference, create a clean and modern portrait grid featuring the same person with several different facial expressions and moods. Arrange the images in a neat 3×3 layout with rounded corners and soft pastel backgrounds. Each portrait should retain the same facial features, hairstyle, and realism, changing only the expression and subtle style. Include expressions such as: natural smile, joyful laugh, calm with eyes closed, neutral and serious gaze, pensive pose with hand on chin, confident and composed expression, sad or disappointed mood, mischievous pout, and surprised reaction with wide eyes. Use soft studio lighting, highly detailed skin texture, shallow depth of field, minimal variations in wardrobe, consistent color palette, ultra-realistic editorial portrait style, high resolution, and clean aesthetics.