mozgi ← 所有提示詞
地鐵監視畫面緊張感十足:一人下行電扶梯,回頭望向鏡頭,旁邊有臉部辨識介面
GPT Image 2.0

地鐵監視畫面緊張感十足:一人下行電扶梯,回頭望向鏡頭,旁邊有臉部辨識介面

提示詞

Create a cinematic surveillance-thriller image set inside a massive multi-level underground metro station. The camera should feel like a ceiling-mounted CCTV positioned high above at roughly a 40-degree angle, looking down over a long industrial escalator and the complex architecture of the station. The composition should use a 4:5 vertical frame, with the escalator cutting diagonally through the image to create strong depth and scale. The station is crowded with hundreds of commuters moving in both directions, forming a chaotic flow of people with subtle motion blur. At the center of the frame, a single person matching the uploaded reference image is descending the escalator. The subject appears to be moving quickly, slightly leaning forward. One hand holds the escalator handrail for balance, while the other carries a small glossy black mini bag with realistic reflections. At a dramatic moment, the subject suddenly turns their head over their shoulder and looks directly into the camera, creating an intense fourth-wall-breaking moment, as if they know they are being watched. Add a realistic CCTV interface overlay, including: Live recording indicators, Timestamp and camera ID, A facial-tracking square locked onto the subject, A secondary zoomed-in facial recognition panel in one corner, Minimal technical HUD elements and surveillance graphics. Lighting should be cold and fluorescent, typical of a modern subway station, with realistic reflections on the metallic escalator surfaces and polished handrails. Keep the surrounding crowd slightly blurred while the subject remains sharp and in focus. Overall mood: high-end documentary surveillance thriller, ultra-photorealistic, cinematic composition, realistic skin textures, natural crowd behavior, detailed environmental reflections, and professional movie-grade realism.
  • 需要你自己的照片:在請求中附上圖片,提示詞會以它為基礎。
在目錄中開啟
此提示詞的權利屬於作者 @frametheory058。任何商業使用(推廣 AI 工具、聚合平台、Telegram 機器人)由你自行負責,mozgi.io 不承擔任何責任。 詳情 →
模型
GPT Image 2.0
比例
4/5
作者
@frametheory058
原文
x.com ↗

相似提示詞

全部提示詞: GPT Image 2.0

已複製