OpenZhenZhen AI プロンプトギャラリー · gpt-image-2.5 · image

レストランで料理を味わう日本人女性の日常スナップ

Candid Shot of a Japanese Woman Tasting Food at a Restaurant

「レストランで料理を味わう日本人女性の日常スナップ」は gpt-image-2.5 を使用した作品です。OpenZhenZhen で元のプロンプトと生成例を確認し、創作の参考にできます。

レストランで料理を味わう日本人女性の日常スナップ — AI image creation
レストランで料理を味わう日本人女性の日常スナップ

プロンプト

中国語プロンプト

创建一张5:8竖幅单图,整齐排列为2x2四格手机抓拍,表现同一名虚构成年日本女性在普通一天中的四个日常瞬间。使用原始标准手机和社交媒体直出质感,细节自然清晰,带轻微JPEG压缩和保留的生活瑕疵;不要影棚拍摄、高级餐厅广告或电影式制作。四格保持人物身份、成年自然比例和日常存在感一致,只改变活动、地点和表情。各格使用符合场景的真实光线,不使用人像模式式过度虚化。 第1格“Morning wake-up bathroom mirror selfie”:小型家用浴室,人物随意站在镜前拿部分可见的手机自拍;镜面构图不完美且略偏,眼睛略浮肿带睡意,短深棕发未打理,素颜或淡妆,宽松随意T恤,浴室顶光柔和。第2格“Outdoor run – mid action selfie”:在户外跑道或人行道上跑步,用前置镜头快速自拍;一只手臂可见,神情专注有活力,张嘴呼吸,穿粉色运动服和白色跑步帽,自然日光,动作感克制。第3格“Coffee break candid”:在小咖啡馆或家中餐桌旁随意坐着喝咖啡,用手持自拍记录;神情若有所思而放松,视线望向别处,咖啡杯正在入口,背景次要安静。第4格“Outdoor breakfast – taken by someone else”:在户外早餐咖啡馆自然进餐,没有察觉镜头,由他人从侧面或桌对面拍摄。 整体保持自然皮肤纹理、真实手部、可信姿态、普通服装和符合场景的光线。早餐格中,从画面左侧伸入一把细长不锈钢勺,接近水平并轻触略偏画面左侧的闭合下唇中央;另一只手支撑脸颊,使其自然轻微受压。画面左下方显示被切断的不锈钢大碗,下方中央偏右放浅金属盘,盘中有赤褐色料理。将准确字符串“@Live_life_style”作为大碗弯曲抛光表面中的微妙变形反射融入照片,而不是后期贴上的水印:文字沿碗面弯曲,被擦痕、雾气、食物或高光打断,明度和对比度低于脸部,只有仔细观察才可辨认。不得漂浮在画面上、独立发光、使用干净数字字体、置于皮肤上或变成签名覆盖层。小型餐饮店背景包含不锈钢设备、奶油色瓷砖、白色百叶设备和带不可辨认细线的模糊告示;不要店名、品牌或其他可读文字。避免未成年感、色情构图、夸张解剖、娃娃脸、塑料皮肤、巨大瞳孔、勉强微笑、勺子进入口腔、华丽配饰、强背景虚化、影棚光、过度整洁的商业布景、App界面、边框、重复人物、多余手指或肢体、手部扭曲、与碗反射无关的水印、标志、字幕和低质量。

英語プロンプト

Create one realistic 5:8 vertical image arranged as a clean 2x2 grid of four candid smartphone photographs of the same fictional adult Japanese woman across one ordinary day. Use raw standard-phone/social-post realism, sharp natural details, slight JPEG compression and preserved imperfections; this is not a studio shoot, luxury restaurant advertisement or cinematic production. Keep her identity, natural adult proportions and everyday presence consistent in all panels, while changing the activity, setting and expression. Use real-world lighting appropriate to each scene and no portrait-mode blur. Panel 1: “Morning wake-up bathroom mirror selfie”—in a small home bathroom, she stands casually before a mirror holding a partly visible phone; imperfect off-center mirror framing, sleepy slightly puffy eyes, unstyled short dark-brown hair, no or minimal makeup, loose random T-shirt and gentle overhead bathroom light. Panel 2: “Outdoor run – mid action selfie”—on a running path or sidewalk, she holds the phone with the front camera while running; one arm visible, focused lively face with open-mouth breathing, pink athletic outfit and white running cap, natural daylight and restrained motion energy. Panel 3: “Coffee break candid”—seated at a small café or home table, she casually drinks from a coffee cup in a handheld selfie, thoughtful relaxed expression looking away, cup mid-sip and unobtrusive background. Panel 4: “Outdoor breakfast – taken by someone else”—at an outdoor breakfast café, she eats naturally and is unaware of the camera, photographed from the side or across the table. For the overall visual language, keep natural skin texture, realistic hands, believable posture, ordinary clothing and scene-specific light. In the breakfast panel, include a narrow stainless-steel spoon entering from the left and touching the center of the closed lower lip slightly toward frame-left; the other hand supports the cheek, causing a slight natural cheek press. Show a cut-off stainless-steel bowl at lower left and a shallow metal plate at lower center-right with reddish-brown food. In the bowl's curved polished reflection, integrate the exact string “@Live_life_style” as a subtle, distorted reflection formed by surrounding light/equipment—not a pasted watermark: it follows the bowl curvature, is interrupted by scratches, fog, food or highlights, lower contrast than the face, and is only recognizable on close inspection. It must not float on the image, glow, use a clean digital font, sit on skin or become a signature overlay. Keep a small eatery background with stainless equipment, cream tiles, a white louvered unit and blurred unrecognizable notices; no store names, brands or other readable text. Avoid underage appearance, sexual framing, exaggerated anatomy, doll-like face, plastic skin, huge pupils, forced smile, spoon inside the mouth, ornate accessories, strong background blur, studio light, over-clean commercial staging, app UI, borders, duplicated people, extra fingers/limbs, distorted hands, watermarks unrelated to the bowl reflection, logos, captions and low quality.

API モデルを見る

画像・動画プロンプトをもっと見る · 画像を生成 · 動画を生成 · AI キャンバスで作成