OpenZhenZhen AI 提示词画廊 · gpt-image-2.5 · image

商场双向扶梯都市时尚抓拍

Urban Fashion Candid on a Mall Escalator

使用人物身份参考创作9:16商场扶梯交错时刻的成年女性都市时尚抓拍，明确摄影者与人物位于相邻反向扶梯。

商场双向扶梯都市时尚抓拍 — AI image creation
商场双向扶梯都市时尚抓拍

提示词 / Prompt

中文提示词

使用上传的人物身份参考图保留明确成年主要女性的面部身份。创作一张真实的9:16竖版都市时尚抓拍：晚间高峰的大型现代商场中，两部并排且方向相反的扶梯交错运行。人物位于画面左侧下行扶梯的较高位置，镜头位于相邻画面右侧上行扶梯的较低位置。清楚表现中央玻璃隔挡、金属结构、两条独立扶手，以及从右侧扶梯斜向左侧拍摄的低位视角。明确成年东亚女性真人Coser自然站在下行扶梯上，不主动面对镜头，一只手搭在扶手上，另一只手拿手机或包，视线望向商场而非镜头。使用类似50–70mm的透视、近距离半身至大腿中部构图、浅景深、真实人流轻微动态模糊、店铺橱窗、灯箱和流动商场氛围。她留自然长发、精致妆容和松弛微微出神的表情，穿不透明合身时尚上衣、宽大但稳妥穿着的都市夹克、短下装配黑色叠层丝袜或短靴，整体为完成度高且非色情化的街头Coser造型。保留自然皮肤与衣料纹理、明亮复杂商场光线和可信空间关系。避免暴露服装、乳房裸露、性化取景、直视镜头、第三人称视角、多余肢体、文字、Logo和水印。

English prompt

Use the uploaded identity reference image to preserve the facial identity of the clearly adult main woman. Create a realistic 9:16 vertical urban-fashion snapshot in a busy modern shopping mall at evening peak hours. Two escalators run side by side in opposite directions: the subject is on the left downward escalator at a higher position, while the camera is on the adjacent right upward escalator at a lower position. Show the central glass partition, metal structure, independent handrails, and the diagonal low-position view from the right escalator toward the left. The adult East Asian coser stands naturally on the descending escalator, not posing for the camera, with one hand on the handrail and the other holding a phone or bag; her gaze turns toward the mall rather than the lens. Use a 50–70mm-like perspective, close half-body to mid-thigh framing, shallow depth of field, realistic passerby motion blur, shop windows, light boxes, signs, and flowing mall atmosphere. Give her long natural hair, refined makeup, a relaxed slightly distant expression, a coordinated opaque fitted fashion top, oversized urban jacket worn loosely but securely, short bottoms with black layered tights or ankle boots, and a polished non-sexual street-cosplay look. Preserve natural skin and fabric texture, bright complex mall lighting, and credible spatial geometry. Avoid revealing clothing, exposed breasts, sexualized framing, direct eye contact, third-person view, extra limbs, text, logos, and watermarks.

查看 API 模型

探索更多图片和视频提示词 · 生成图片 · 生成视频 · 在 AI 画布中创作