PROMPT RECORD图像记录

https://t.co/14FmgpR1QV 仅参考已上传的肖像进行身份匹配：面部结构、脸型、发型与发色、肤色以及真实的人类外观。切勿复制任何外部参考中的面部。以手机拍摄的方式，将主体重新构...

GPT-Image-2 iamaiistudio Tue Jun 09 08:07:54 +0000 2026

https://t.co/14FmgpR1QV

prompt:

Reference the uploaded portrait solely for identity matching: facial structure, face shape, hair style and color, skin tone, and authentic human likeness. Never duplicate the face from any external reference. Reimagine the subject in a photorealistic overhead close-up portrait with a cute maid aesthetic captured on a phone.

Position the camera angled downward from above, very near the face, using a phone-camera wide perspective. The subject gazes straight up into the lens with a serene, gentle, faintly enigmatic look. Maintain strong facial recognition, preserving natural bone structure, real skin texture, authentic eyes, true hair detail, no artificial or doll-like appearance.

Setting and wardrobe: dimly lit interior, intense cool blue-white phone flash, slightly blown-out face highlights, cinematic teal color grade, dense shadows, softened blur along the frame edges. Dress the subject in a charming pastel maid-inspired ensemble: rounded white collar, white ruffled apron, delicate pink accents, pink polka-dot cap or hair ornament, modest and tasteful. One hand holds a small pink kitchen spatula at the frame's edge. Tight head-and-shoulders framing, narrow depth of field, authentic low-light candid quality, 9:16 vertical format.

Negative Prompt:

watermark, logo, text, Chinese characters, anime, cartoon, CGI, plastic skin, doll face, over-smoothed skin, fake face, distorted eyes, crossed eyes, bad anatomy, extra fingers, deformed hands, oversized head, low quality, blurry face, harsh beauty filter, nudity, overly revealing outfit, duplicate person, AI artifacts.

#AIart #GPTImage2

https://t.co/14FmgpR1QV 仅参考已上传的肖像进行身份匹配：面部结构、脸型、发型与发色、肤色以及真实的人类外观。切勿复制任何外部参考中的面部。以手机拍摄的方式，将主体重新构想为具有可爱女仆美学的逼真俯视特写肖像。 将相机从上方斜向下放置，非常靠近面部，使用手机摄像头的广角透视。主体直接向上凝视镜头，神态宁静、温柔、略带神秘。保持强烈的面部识别度，保留自然骨骼结构、真实皮肤纹理、真实的眼睛、真实的头发细节，避免人工或玩偶般的外观。 场景与服装：昏暗的室内环境，强烈冷蓝白色手机闪光灯，面部高光略微过曝，电影感青绿色调，浓重阴影，画面边缘柔和模糊。为主体穿上迷人的粉彩女仆风格服装：圆形白领、白色荷叶边围裙、精致粉色点缀、粉色波点头巾或发饰，端庄而雅致。一只手在画面边缘握着一把粉色小厨房铲。紧凑的头...

查看来源

中文说明

https://t.co/14FmgpR1QV 仅参考已上传的肖像进行身份匹配：面部结构、脸型、发型与发色、肤色以及真实的人类外观。切勿复制任何外部参考中的面部。以手机拍摄的方式，将主体重新构想为具有可爱女仆美学的逼真俯视特写肖像。将相机从上方斜向下放置，非常靠近面部，使用手机摄像头的广角透视。主体直接向上凝视镜头，神态宁静、温柔、略带神秘。保持强烈的面部识别度，保留自然骨骼结构、真实皮肤纹理、真实的眼睛、真实的头发细节，避免人工或玩偶般的外观。场景与服装：昏暗的室内环境，强烈冷蓝白色手机闪光灯，面部高光略微过曝，电影感青绿色调，浓重阴影，画面边缘柔和模糊。为主体穿上迷人的粉彩女仆风格服装：圆形白领、白色荷叶边围裙、精致粉色点缀、粉色波点头巾或发饰，端庄而雅致。一只手在画面边缘握着一把粉色小厨房铲。紧凑的头...

原始 Prompt

https://t.co/14FmgpR1QV

prompt:

Reference the uploaded portrait solely for identity matching: facial structure, face shape, hair style and color, skin tone, and authentic human likeness. Never duplicate the face from any external reference. Reimagine the subject in a photorealistic overhead close-up portrait with a cute maid aesthetic captured on a phone.

Position the camera angled downward from above, very near the face, using a phone-camera wide perspective. The subject gazes straight up into the lens with a serene, gentle, faintly enigmatic look. Maintain strong facial recognition, preserving natural bone structure, real skin texture, authentic eyes, true hair detail, no artificial or doll-like appearance.

Setting and wardrobe: dimly lit interior, intense cool blue-white phone flash, slightly blown-out face highlights, cinematic teal color grade, dense shadows, softened blur along the frame edges. Dress the subject in a charming pastel maid-inspired ensemble: rounded white collar, white ruffled apron, delicate pink accents, pink polka-dot cap or hair ornament, modest and tasteful. One hand holds a small pink kitchen spatula at the frame's edge. Tight head-and-shoulders framing, narrow depth of field, authentic low-light candid quality, 9:16 vertical format.

Negative Prompt:

watermark, logo, text, Chinese characters, anime, cartoon, CGI, plastic skin, doll face, over-smoothed skin, fake face, distorted eyes, crossed eyes, bad anatomy, extra fingers, deformed hands, oversized head, low quality, blurry face, harsh beauty filter, nudity, overly revealing outfit, duplicate person, AI artifacts.

#AIart #GPTImage2

浅景深粉彩极近特写俯视头饰或帽子梦幻需要复核人物肖像