OpenZhenZhen AI 프롬프트 갤러리 · gpt-image-2.5 · image

3:4 사실적과 일러스트 분할 인물화

Realistic and Illustrated Split Portrait 3:4

‘3:4 사실적과 일러스트 분할 인물화’은(는) gpt-image-2.5로 만든 작품입니다. OpenZhenZhen에서 원본 프롬프트와 생성 예시를 살펴보고 새로운 창작에 활용해 보세요.

3:4 사실적과 일러스트 분할 인물화 — AI image creation
3:4 사실적과 일러스트 분할 인물화

프롬프트

중국어 프롬프트

使用附带的 ref-10a-v1 照片作为唯一来源和身份参考,创作一张高细节竖版 3:4 分割人像,画面中始终是同一位成年女性。由于参考图较宽,只进行形成 3:4 画幅所需的居中裁取;完整保留她的头肩姿势、面部位置、服装轮廓和周围可见背景,不拉伸图像,也不改变姿势。用一条居中的竖向撕纸边缘或克制的故障风过渡,将人物处理分成左右相等两半。左侧:人物保持写实,保留原始光线、纹理和精细细节,只轻微提升清晰度,并在人物周围加细微的柔和粉彩色或霓虹手绘涂鸦轮廓。右侧:只将同一位女性转化为可爱的现代数字插画,采用干净线稿、柔和阴影、粉彩或鲜艳色彩以及精致的 Q 版到半写实处理;严格保持相同面部、年龄感、表情、姿势、相机角度、服装轮廓和位置。可以在插画人物附近点缀少量贴纸风涂鸦或可爱图标,但不要遮挡人物。原有青蓝色背景在左右两侧都必须保持视觉上不变:保留相同色彩、纹理、光线和透视;不要重绘、替换、扩展成其他场景,也不要风格化任何背景区域。保持分割线两侧连续。追求超细节、清晰的 8K 级视觉质感,这只是质量目标,不是可见文字,也不代表实际输出尺寸。构图保持干净,适合社交媒体。不要增加人物、改变背景、让表情不一致、遮挡主体、添加水印、精确字样或标志。

영어 프롬프트

Use the attached ref-10a-v1 photo as the sole source and identity reference. Create one high-detail vertical 3:4 split-style portrait of the same single adult woman. Because the reference is wide, use only a centered crop needed to form the 3:4 frame; keep her complete head-and-shoulders pose, facial placement, clothing contours and the surrounding background visible, without stretching or changing her pose. Divide the woman's treatment into equal left and right halves with one centered vertical torn-paper edge or restrained glitch transition. LEFT: keep the woman photorealistic, preserving the original light, texture and fine detail, with only a slight clarity lift and a subtle soft-pastel or neon hand-drawn doodle outline around her. RIGHT: transform only the same woman into a cute modern digital illustration with clean line art, soft shading, pastel or vibrant color and a polished chibi-to-semi-realistic treatment; preserve exactly the same face, age impression, expression, pose, camera angle, clothing silhouette and placement. A few small sticker-style doodles or cute icons may sit near the illustrated subject without covering her. The original teal-blue background must remain visually unchanged on both sides: keep the same color, texture, lighting and perspective; do not redraw, replace, extend with a different scene, or stylize any background area. Preserve continuity across the seam. Aim for an ultra-detailed, crisp 8K-quality visual finish as a quality aspiration, not as visible text or a claim about the actual output dimensions. Keep the composition clean and suitable for social media. Do not add people, change the background, mismatch the expression, obscure the subject, add watermarks, or introduce exact text or logos.

API 모델 보기

더 많은 이미지·동영상 프롬프트 보기 · 이미지 생성 · 동영상 생성 · AI 캔버스에서 만들기