聊天GPT

像一本書封面

為聊天GPT 編寫。不同型號的結果和參考照片支援可能有所不同。

使用提示詞前

列出的工具:聊天GPT。完整說明:4,516 個字元。 1 個提供的範例影像。

This prompt mentions reference images. Read its instructions for the required number of photos and identity constraints. We have not independently tested this prompt. Tool labels and supplied images are not a guarantee of compatibility or output quality.

完整提示詞

Create an ultra-high-quality, highly accurate image using the uploaded reference image as the ABSOLUTE STRICT VISUAL AND COMPOSITION REFERENCE.

REFERENCE IMAGE RULE:
Recreate the reference image as closely and accurately as technically possible. The reference image is the ONLY source for:

overall composition

character count

character positions

body orientation

poses

head angles

facial expressions

eye direction and gaze

body proportions and figure

clothing and accessories

hairstyle styling

lighting and shadows

color palette

background

framing and crop

scale and proportions of every character

overall artistic style and visual treatment


Do not redesign, rearrange, simplify, or reinterpret the reference image.

IDENTITY REPLACEMENT RULE — EXTREMELY IMPORTANT:
Replace ONLY the facial identities of the characters with the uploaded face references.

For EVERY MALE CHARACTER:
Use my uploaded male face photo as the ONLY source of facial identity. All male characters must use my exact recognizable facial identity, while preserving their individual original expressions, head angles, and poses from the reference image.

For EVERY FEMALE CHARACTER:
Use my uploaded female face photo as the ONLY source of facial identity. All female characters must use my exact recognizable female facial identity, while preserving their individual original expressions, head angles, and poses from the reference image.

Do not invent new faces. Do not mix identities. Do not use the original characters’ facial identities. Do not use any other person’s face.

FACIAL IDENTITY PRESERVATION:
Preserve the supplied face references’ recognizable facial structure, face shape, forehead, eyebrows, eyes, nose, lips, jawline, chin, skin tone, and natural facial proportions. The final faces must clearly look like the people in the uploaded identity references.

Do not beautify, redesign, reshape, slim, widen, age, masculinize, feminize, or otherwise alter the supplied facial identities.

EXPRESSION AND POSE RULE:
The REFERENCE IMAGE is the ONLY source for every character’s expression.

Preserve the exact expression of each corresponding character:

same emotional intensity

same eye direction

same eyelid position

same eyebrow position

same mouth shape

same lip position

same jaw tension

same head tilt

same facial angle

same overall attitude


The supplied face identity must be fitted into the reference character’s expression, NOT the other way around.

Do not give every character the same expression. Each character must retain the exact expression and visual attitude of the corresponding reference character.

STYLE AND VISUAL TREATMENT:
Preserve the reference image’s exact artistic style, including its level of realism or stylization, facial rendering, edge treatment, shadows, highlights, color blocking, texture, contrast, and overall visual language.

Do not turn the image into a different art style. Do not add unnecessary photorealistic details if the reference is stylized. Do not make the image cartoonish if the reference is realistic.

For this specific reference, preserve the bold graphic portrait aesthetic:

dramatic red, black, white, yellow, and dark blue-gray color palette

high-contrast graphic color blocking

sharp, expressive painted edges

strong shadow shapes

stylized but recognizable facial rendering

bold poster-like composition

textured artistic finish

no unnecessary text or typography


COMPOSITION LOCK:
Keep the exact same:

number of characters

character placement

body position

pose

head direction

camera angle

framing

crop

scale

background

lighting

color arrangement


Do not move, add, remove, duplicate, or rearrange any character.

FINAL RESULT:
The final image must look like the SAME original reference artwork, recreated with the supplied facial identities.

The viewer should immediately recognize:

1. The original reference composition and visual style.


2. The original characters’ expressions and poses.


3. The supplied male and female facial identities.



Only the faces/identities should change. Everything else must remain as close to the reference image as technically possible.

QUALITY CONTROL:
Before finalizing, compare the generated image against the reference image and correct any differences in composition, character placement, pose, expression, head angle, framing, colors, and artistic style.

Do not add text, logos, watermarks, captions, or extra elements unless they already exist in the reference image.

複製完整的提示並根據要求使用您自己的參考照片。

打開ChatGPT ↗

此提示提供的範例影像。此提示尚未經過我們網站的測試。

照相館
AI 產生的範例

為您自己的照片提供一點靈感。

1加入您的照片
A 人

保持臉部清晰可見

B 人

保持臉部清晰可見

JPG、PNG 或 WebP · 最大。每個 10 MB

僅當您選擇「產生」時才會發送您的照片。使用您有權使用的照片。

2讓它成為你的
更改詳情
編輯您的提示
開放雙子座

準備好…

登入您的帳戶

繼續使用您的 Google 帳戶。

您無需登入即可瀏覽和複製提示。

我們如何使用您的資訊: 隱私權政策