现实在前,混乱在后。

GPT-Image-2 · @MahnoorAi12 · Thu Jul 30

在画廊中打开

分享链接 · https://openimages.ajiang.me/p/2082708020988154179/

现实在前,混乱在后。 ***""""GPT image 2 提示词 """"**** 仅使用上传的参考图像作为唯一的面部参考,并100%完整保留面部身份,包括脸型、眼睛、眉毛、鼻…
camera_lens · 85mm-portraitcamera_lens · shallow-depth-of-fieldcomposition · full-bodyhuman_attributes · long-hairlighting · studio-lightmood · playful

提示词(原文,逐字保留)

Reality in front, chaos behind. ***""""GPT image 2 prompt """"**** Use the uploaded reference image as the ONLY face reference and preserve 100% exact facial identity, including the face shape, eyes, eyebrows, nose, lips, jawline, skin tone, facial proportions, expression, age, and every unique facial feature with STRICT IDENTITY LOCK. Do not copy or reference the hairstyle, hairline, or hair volume from the uploaded image. Generate the hairstyle exactly as described below, independent of the reference image. Create an ultra-realistic editorial fashion portrait of the same young woman standing calmly. Give her a stylish medium-length wavy hairstyle with soft, natural volume. She is wearing an oversized trendy shirt, relaxed-fit cargo pants, and fashionable sneakers. Her expression is cute and slightly confused, with a subtle head tilt that enhances the playful mood. Behind her, create a playful hand-drawn doodle shadow that mimics her silhouette in exaggerated, funny poses, adding a humorous artistic contrast while keeping the real subject perfectly realistic and sharply in focus. Use cinematic lighting, ultra-realistic photography, high contrast, premium editorial fashion styling, realistic skin texture, detailed fabric folds, clean minimal composition, shallow depth of field, an 85mm lens, natural color grading, 8K ultra-HD quality, and a 4:5 aspect ratio. STRICT IDENTITY LOCK: Do not alter, beautify, stylize, or redesign the face in any way. Preserve the exact facial identity from the uploaded reference image with complete facial consistency while using only the newly described hairstyle, not the hairstyle from the uploaded reference.

中文译文

现实在前,混乱在后。 ***""""GPT image 2 提示词 """"**** 仅使用上传的参考图像作为唯一的面部参考,并100%完整保留面部身份,包括脸型、眼睛、眉毛、鼻子、嘴唇、下颌线、肤色、面部比例、表情、年龄以及每一个独特的面部特征,使用严格身份锁定(STRICT IDENTITY LOCK)。不要复制或参考上传图像中的发型、发际线或发量。完全按照下面的描述独立生成发型,与参考图像无关。 为同一位年轻女性创作一张超现实的编辑时尚肖像,她平静地站立。给她一个时尚的及肩中长波浪发型,带有柔和自然的丰盈感。她身穿一件宽松的潮流衬衫、宽松版型的工装裤和一双时尚的运动鞋。她的表情可爱且略带困惑,头部有一个轻微的倾斜,强化俏皮的氛围。 在她身后,创建一个俏皮的手绘涂鸦影子,模仿她的轮廓,做着夸张滑稽的姿势,在保持真实主体完美逼真且对焦清晰的同时,增加一种幽默的艺术对比。 使用电影级光线、超现实摄影、高对比度、顶级编辑时尚造型、真实的皮肤纹理、精细的织物褶皱、简洁极简的构图、浅景深、85mm镜头、自然色彩调色、8K超高清画质,以及4:5的宽高比。 严格身份锁定(STRICT IDENTITY LOCK):不要以任何方式改变、美化、风格化或重新设计面部。完整保留上传参考图像中的确切面部身份,保持完全的面部一致性,同时仅使用新描述的发型,而非上传参考图像中的发型。

来源与署名

原文由 @MahnoorAi12 发布在 X:查看原推文。 本页逐字保留原文并提供机器翻译的中文解读;版权归原作者所有。

在画廊中浏览

同工具的更多提示词