AI-фотоNano Banana 2

Уличный вайб — реалистичное селфи (543)

Промпт для реалистичное селфи: уличный вайб, сохранение лица по исходному фото и готовая структура для быстрой генерации.

Заказать адаптацию
Модель: Nano Banana 2Формат: 3:4Сложность: Легко
Image 15973

Промпт

Nano Banana 2

Use uploaded selfie as ONLY face reference; preserve exact likeness, exact age, facial proportions, hair color and hair length, natural skin texture, identity preservation, no face mixing. Ultra-realistic cinematic night photo at a gas station in the rain, same framing and body language: medium shot, subject standing in side profile with the body turned away from the camera and the head turned back toward the camera, shoulders slightly slouched, arms relaxed low, calm detached expression. Keep the composition tight and vertical, with the subject placed close to the foreground and the gas station structures filling the background. Location: modern gas station forecourt with wet reflective pavement, fuel pumps visible, slightly different layout from usual, darker and emptier atmosphere, and a black BMW E34 subtly visible in the background near the pumps, only partially seen and softly out of focus. Outfit: if female — oversized dark bomber jacket or oversized washed leather jacket over a fitted light cropped top, dark low-rise jeans; if male — oversized dark bomber jacket or oversized washed leather jacket over a fitted light t-shirt or ribbed top, dark jeans. Hair looks slightly wet and messy from the rain. Keep the exact color grading: strong green-cyan cast in the shadows and skin, deep teal-blue night sky, warm yellow-orange gas station lights, wet asphalt reflecting green and gold tones, moody high-contrast neon realism. Soft bloom around lights, subtle film grain, realistic skin detail, premium editorial realism. Camera: 50mm, f/1.8, ISO 800

Как использовать промпт

01

Загрузите фото лица

Выберите исходное фото, где лицо хорошо видно и подходит под нужный ракурс.
02

Вставьте промпт

Скопируйте промпт и запустите генерацию в выбранной нейросети.