OpenZhenZhen AI プロンプトギャラリー · GPT Image 2.6 · image

ミニマルな浴室ミラーセルフィー

Minimalist Bathroom Mirror Selfie

「ミニマルな浴室ミラーセルフィー」は GPT Image 2.6 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。

ミニマルな浴室ミラーセルフィー — AI image creation
ミニマルな浴室ミラーセルフィー

プロンプト

中国語プロンプト

创作一张写实的浴室镜面自拍肖像。主体为实际完整的 identity_reference ref-10a-v1 中的同一位成年女性,该参考只使用一次,作为唯一身份参考。保留参考图可辨认的面部特征、自然肤色和深棕色长发。只匹配参考中实际可见的身份与外观;不要声称像素级完全复制,也不要臆造参考中看不到的身材比例。 她坐在浴室台面边缘,身体略微侧转,越过肩膀望向镜子,并单手举起手机自拍。呈现与人物和手机相符、方向合理的真实镜面反射;镜中是同一位女性,不要重复生成另一人物。肩膀和姿态自然放松。 来源未指定服装。明确选择全不透的炭灰色短袖上衣和深色全覆盖长裤作为服装默认,版型整洁随性,整体自然得体。 场景设在现代极简浴室,包含浅色中性墙面、玻璃淋浴间、整齐叠放的毛巾和大理石台面。采用柔和自然的浴室光线、平衡曝光与轻柔阴影。画面应像真实手机镜面自拍,反射准确,面部细节清晰,皮肤纹理自然,构图干净可信。来源未指定整体画幅;明确采用竖版 4:5 作为肖像构图默认。

英語プロンプト

Create a photorealistic mirror-selfie portrait of the same adult woman shown in the actual full identity_reference ref-10a-v1, used exactly once as the sole identity reference. Preserve her recognizable reference-visible facial features, natural complexion and long dark-brown hair. Match only appearance and details visible in that reference; do not claim pixel-perfect replication or invent unseen body proportions. She is sitting on the edge of a bathroom counter, her body turned slightly to the side as she looks over her shoulder toward the mirror. She raises a phone in one hand to take the selfie. Show one coherent mirror reflection of this same woman and phone, with physically plausible reflection direction and no duplicated person. Her shoulders and posture remain relaxed and natural. The source does not specify clothing. As the disclosed outfit default, dress her in a fully opaque charcoal-gray short-sleeved top and dark full-coverage trousers, with a neat, casual fit. Keep the styling modest and natural. Set the scene in a modern minimalist bathroom with light neutral walls, a glass shower enclosure, neatly folded towels and a marble countertop. Use soft natural bathroom light, balanced exposure and gentle shadows. Make it feel like a realistic phone mirror selfie, with accurate reflections, sharp facial detail, natural skin texture and a clean, believable composition. The source does not specify an overall canvas ratio; use vertical 4:5 as the clearly disclosed portrait-framing default.

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