OpenZhenZhen AI 提示词画廊 · gpt-image-2.5 · image
超市爱心框景治愈系少女日常
Heart Frame in the Supermarket: A Healing Girl's Daily Life
保留实际成年女性的面部身份、比例与发型，创作9:16超市抓拍，呈现购物车、匿名前景手心形框、浅灰卫衣及复古漏光。 Create a 9:16 supermarket candid preserving the actual adult woman’s identity, proportions and hair, with a cart, anonymous heart-framing hand, gray sweatshirt and nostalgic light leak.
提示词 / Prompt
中文提示词
仅将实际 ref-10a-v1 肖像用作这位成年女性的面部身份、脸型、自然特征、肤色、身材比例和发型参考。不要复制该肖像的衣服、背景、姿势、光线或图像质量。创作一张写实真人抓拍照片，场景为超市过道，垂直比例 9:16。同一位成年女性以相同可辨认的面容和发型自然倚靠在购物车上，带着温柔平静的表情向上看零食货架。前景中一只匿名的大手虚化成爱心框景；只出现这只手，不出现其他人物或面孔。她穿宽松的浅灰色卫衣，造型简约休闲，头发自然散落，皮肤纹理真实。采用柔和室内荧光灯、胶片相机颗粒感和轻微动态模糊；右上角加入温暖的橙色漏光，营造怀旧日常抓拍与真实生活氛围。不加水印、文字或标志。反向提示：水印、标志、文字、AI痕迹、动漫、卡通、插画、塑料感皮肤、过度平滑的皮肤、假脸、改变身份、不同发型、不良解剖结构、变形的手、多余手指、模糊的脸、糟糕的眼睛、扭曲的购物车、混乱构图、低质量、脸部过曝、不真实的光线。
English prompt
Use the actual ref-10a-v1 portrait only as the reference for the adult woman’s facial identity, face shape, natural features, skin tone, body proportions and hairstyle. Do not copy that portrait’s clothing, background, pose, lighting or image quality. Create one realistic live-action candid photo in a supermarket aisle, vertical 9:16. The same adult woman, with the same recognizable face and hairstyle, leans naturally on a shopping cart and looks upward toward the snack shelves with a soft, calm expression. In the foreground, an anonymous large blurred hand forms a heart frame around her; show only the hand, with no other person or face. Give her an oversized light-gray sweatshirt and simple casual styling, with natural loose hair and realistic skin texture. Use soft indoor fluorescent lighting, film-camera grain, slight motion blur and a warm orange light leak in the upper-right corner for a nostalgic everyday snapshot with an authentic real-life atmosphere. No watermark, text or logo. Negative prompt: watermark, logo, text, AI mark, anime, cartoon, illustration, plastic skin, over-smoothed skin, fake face, changed identity, different hairstyle, bad anatomy, deformed hands, extra fingers, blurry face, bad eyes, distorted shopping cart, messy composition, low quality, overexposed face, unrealistic lighting.
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
相关作品
- Mia OOTD 贴纸标注
A restrained bilingual OOTD sticker-note overlay for the actual full portrait, labeling only the visible black camisole.
- 夜幕下的复古面包车与黑发女性
A complete bilingual JSON prompt template for an analog night portrait with a vintage VW Combi and a white board; each source example is resolved to a concrete value.
- 夜晚斑马线上的酷飒街头肖像
Create a cinematic, photorealistic nighttime street portrait with a relaxed, confident seated pose, specified streetwear, warm streetlights and shallow depth of field, while preserving the configured identity reference exactly.
- 俏皮手绘涂鸦互动照片
保留整张参考肖像的主体、构图与光线，在发丝和边缘加入轻量互动涂鸦及一句情境短文案。 Preserve the full reference portrait, composition and lighting, adding light interactive doodles around the hair and edges with one contextual short note.
- 欢快萌系皮克斯风格3D卡通人像
将实际参考女性转为可爱3D动画肖像，保留原脸、发型、独特表情与原服装，置于纯白影棚背景。 Turn the actual reference woman into a cute 3D animated portrait while preserving her face, hairstyle, unique expression and outfit against a plain white studio background.
- 同一角色八姿势3D萌系贴纸表
以实际成年女性参考为身份依据，制作同一角色八种情绪动作的3D贴纸表，统一服装与白色描边。 Create a 3D sticker sheet of the same reference adult woman in eight emotional poses, with consistent clothing and white outlines.