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

日当たりの良い窓辺で自撮りする東アジアの女性

East Asian Woman Taking a Selfie by a Sunny Window

「日当たりの良い窓辺で自撮りする東アジアの女性」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。

日当たりの良い窓辺で自撮りする東アジアの女性 — AI image creation
日当たりの良い窓辺で自撮りする東アジアの女性

プロンプト

中国語プロンプト

使用用户上传的参考图作为画面中主要可见的东亚成年女性的面部身份依据。保持可识别的面部结构、真实肤色、细微不对称和未经修饰的自然特征,不使用美颜滤镜或风格化渲染。创作一张超写实的随性手机自拍:安静房间中,成年女性坐在阳光照射的窗边。以近距离手臂长度的放松三分之二侧面拍摄,她略微转向窗户并凝视窗外,表情深思而含蓄,眼神平静,面部肌肉放松,嘴部自然闭合。她拥有深棕黑色长发,自然蓬松、略显凌乱,逆光中可见碎发和单根发丝。她戴大型圆形黑色醋酸纤维眼镜,镜框可信地反射日光,穿合身麻灰色棉质T恤,表现自然褶皱和垂坠。使用从一侧进入的自然阳光,形成真实对比与横跨脸部和上半身局部的窗框阴影带;保留真实高光和色彩变化,不要HDR或人工光晕。背景保持简约:浅中性色墙壁、干净的垂直窗框,不加入装饰干扰。使用略低的手机前置摄像头角度,前景部分可见手臂,带轻微广角透视、真实计算摄影、细微数字噪点、轻微镜头柔和感和自然曝光。紧凑上半身竖幅构图,主体略偏离中心,窗户位于画面边缘,保留真实手持自拍的不完美感。优先保证准确解剖结构、皮肤毛孔、自然发丝、可信光学瑕疵和纪录片式手机画质;不要CGI感、电影调色、网红式修图、动漫、插画、标志、水印或文字。

英語プロンプト

Use the uploaded reference image as the facial identity reference for the main visible adult East Asian woman. Preserve recognizable facial structure, true-to-life skin tone, subtle asymmetry, and authentic unretouched features without beautification filters or stylized rendering. Create an ultra-photorealistic casual smartphone selfie in a quiet room beside a sunlit window. Capture a relaxed three-quarter profile from close arm's length as she turns slightly toward the window and gazes outside with a thoughtful understated expression, calm eyes, relaxed facial muscles, and a naturally resting mouth. Give her long dark brown-black hair with natural volume, slight disarray, flyaways, and individual strands catching the backlight. She wears large circular black acetate eyeglasses with believable daylight reflections and a simple fitted heather-gray cotton T-shirt with natural creasing and drape. Use one-sided natural sunlight with convincing contrast and window-frame shadow bands across parts of the face and upper body; retain realistic highlights and color variation without HDR or artificial glow. Keep the setting minimal: light neutral wall, clean vertical window frame, no decorative distractions. Use a slightly low front-camera angle, partially visible arm, mild wide-angle distortion, realistic computational photography, subtle digital noise, slight lens softness, and natural exposure. Compose a tight upper-body vertical frame with the subject slightly off-center and the window at the edge, preserving genuine handheld imperfection. Prioritize accurate anatomy, skin pores, natural hair, believable optics, documentary smartphone quality, and no CGI, cinematic grading, influencer retouching, anime, illustration, logos, watermarks, or text.

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