一、核心目标与强约束
1. 精准姿态复刻:以图 2 为姿态的唯一参考,对图 1 中人物的全身姿态进行“点对点对齐”复刻(确保头部、颈部、肩部、胸部、臀部、膝盖、脚踝、手臂、手腕、手指的朝向和弯曲程度一致)。
2. 运镜一致:确保相机位置、拍摄角度、镜头高度、俯仰角、偏航角、透视关系、景深效果与图 2 一致。
3. 构图一致:确保人物在画面中的位置、占比、负空间比例、裁剪方式(脚/头/手是否被裁切)与图 2 一致。
4. 头部与视线一致:确保头部倾斜角度、下巴抬/收程度、视线方向与图 2 一致。
5. 外观禁止:严格保留图 1 人物身份特征与风格(脸型、五官、发型、发色、肤色、服装款式与颜色、材质、配饰、纹理、身材比例、年龄、气质、艺术风格/渲染风格),仅改变姿态与运镜构图以匹配图 2。

二、执行方法与校准要求
1. 对齐方式:以图 2 中人物的骨骼为“目标骨骼”,将图 1 中人物的骨骼“重定向”到目标骨骼;肢体角度、重心位置、躯干扭转幅度必须匹配。
2. 关键点优先级:优先匹配重心与骨盆朝向,其次匹配躯干旋转与肩线,再次匹配四肢(手、脚),最后微调手指与表情细节。
3. 允许误差范围:姿态与运镜误差应尽可能接近 0;如不可避免,误差必须最小且不应在视觉上显而易见。
4. 禁止事项:禁止增加或删除物体;禁止改变服装款式与细节;禁止改变人物风格;禁止改变背景风格导致构图变化;禁止将图 2 人物的外观特征迁移到图 1 人物上。
产品营销 - Nano Banana Pro 的高保真姿态复制提示
0
0
0

一个高度受限的多点提示,专为 Nano Banana Pro 设计,用于将参考图像(图 2)中的复杂姿势精确复制到源角色(图 1)上。它强制严格对齐身体点、摄像机角度和构图,并保留原始角色的身份和风格,使其成为复制高难度动作或时尚摄影的理想选择。

Language
EnglishDeutschEspañolFrançais日本語한국어Português中文
Created7 months ago
Last updated7 months ago
Creator

Services with a clipboard icon will copy the prompt to your clipboard first.

Version History
Prompt documentation
Comments (0)
Please log in to leave a comment.

Be the first to comment

to start the conversation.

Related Prompts
EnglishDeutschEspañolFrançais日本語한국어Português中文

An extremely detailed, structured JSON prompt for generating a beach portrait with a 'strict visual replication' mode, focusing intensely on the subject's pose, body geometry, and surface details (wet skin sheen, natural compression, genuine smile). The prompt meticulously describes the posture, limb positions, and environment for maximum fidelity to a reference image.

0
0
0

EnglishDeutschEspañolFrançais日本語한국어Português中文

A refined JSON prompt, based on a previous one, that enforces strict replication of the subject's facial features, hairstyle, and inner wear from a reference image, while specifically retaining an oversized electric pink wool trench coat. It aims for a high-fashion editorial look with contrasting colors and spotlighting.

0
0
1

EnglishDeutschEspañolFrançais日本語한국어Português中文

A prompt designed for image-to-image generation, requiring the AI to replicate an uploaded clothing item exactly across two framed images: one as a vintage kawaii scrapbook illustration, and the other as a photorealistic photo of an adult Asian woman wearing the outfit in a cozy cafe setting.

0
0
0

EnglishDeutschEspañolFrançais日本語한국어Português中文

A prompt specifically tailored for e-commerce product listings, focusing on generating a full-body image of a model wearing a highly detailed, embroidered dark pinstripe blazer. The prompt ensures exact replication of the embroidery patterns from a reference image and sets the scene in an elegant, softly lit modern living room, prioritizing texture and accurate color representation.

0
0
0

EnglishDeutschEspañolFrançais日本語한국어Português中文

An extremely detailed, structured prompt for Nano Banana Pro focused on replicating a specific surreal fashion concept from a reference image. The core concept is an illusion where an open magazine held in front of the face perfectly replaces the subject's face with the printed model's face, including a seamless 'hair extension' column. It mandates strict identity lock and ultra-photoreal editorial style.

0
0
0

EnglishDeutschEspañolFrançais日本語한국어Português中文

An extremely technical prompt for generating a hyper-realistic mirror selfie, simulating an iPhone 17 Pro capture. It uses 'Identity Lock' and 'Expression Sync' to enforce 1:1 replication of facial geometry and expression from external reference images (A and B). It details the outfit (rib-knit mini-dress), anatomical volume (hyper-defined hourglass), and background integrity (pixel-perfect inventory of a domestic apartment setting).

0
0
0