ai21.net
语言

Hand-drawn Doodle Art Overlay

原帖未提供效果图

提示词

://t.co/jBGrThiRdq

prompt:

Take the uploaded photo and keep the subject completely unchanged — same face, pose, framing, and lighting, no warping or distortion. Layer expressive hand-drawn illustrations on top that playfully interact with the person or scene.

Have the doodles react to what's in the frame: follow gestures, amplify movement, trace body outlines, or bring in imaginative details that feel connected to the moment. The sketches should look deliberate, like someone drew straight onto the print with spontaneous yet thoughtful pen strokes.

Go for a loose, slightly rough drawing style — wobbly lines, organic curves, a sketchbook aesthetic. Add handwritten words or short captions that fit the vibe of the image: make them clever, context-specific, and genuinely playful rather than generic filler.

Keep the illustrated additions from overwhelming the original photo. Aim for a balanced final composition with crisp, high-res output and a color palette that feels native to the original scene.

#AIart #GPTImage2

中文译文

://t.co/jBGrThiRdq

提示词:

取上传的照片,保持主体完全不变——同一张脸、同样的姿态、构图与光线,不得变形或扭曲。在其上叠加富有表现力的手绘插画,与人或场景进行俏皮的互动。

让涂鸦对画面中的内容作出反应:跟随手势、放大动作、勾勒身体轮廓,或加入与此刻感觉相通的富有想象力的细节。这些素描应显得是刻意画上去的,就像有人用随性却又经过思考的笔触直接画在照片上。

采用松散、略显粗糙的绘画风格——抖动的线条、有机的曲线、速写本的美感。添加契合画面氛围的手写文字或简短说明:让它们机巧、贴合语境、真正俏皮,而不是泛泛的填充内容。

不要让插画元素压过原始照片。追求平衡的最终构图,清晰、高分辨率的输出,以及与原始场景自然契合的配色方案。

#AIart #GPTImage2

溯源

原作者
iamaiistudio
发布日期
2026-06-09(由推文 ID 反解)
原始出处
查看原帖 ↗
收录来源
EvoLinkAI/awesome-gpt-image-2-API-and-Prompts · 2064454454553051298 · CC0-1.0