REFERENCE:
Use the uploaded reference photo as the exact identity reference for the person.

DOG: [تعداد و نوع سگ رو اینجا بزنید]

Create a highly photorealistic 9:16 lifestyle street photograph inspired by the provided scene.

IDENTITY — HIGHEST PRIORITY:
Preserve the exact identity of the person from the uploaded reference photo with maximum accuracy, including facial structure, proportions, skin tone, eyes, eyebrows, nose, lips, jawline, age, hair, and all distinctive facial features. The person must remain instantly recognizable. Do not beautify, reshape, reinterpret, or alter the face.

SCENE:
The person is naturally walking on a clean, sunlit urban sidewalk while walking the specified dog(s) on realistic leashes. Include trees, storefronts, buildings, parked cars, and only a few distant pedestrians. Keep the sidewalk relatively uncrowded.

The person holds a smartphone in one hand and casually looks at its screen while walking. In the other hand, they hold a takeaway coffee cup and the dog leash(es). The pose and movement must feel candid and natural.

OUTFIT:
Automatically choose a stylish, modern casual outfit appropriate for the person in the uploaded reference photo. Adapt the clothing naturally based on whether the reference person is female or male. Do not copy a fixed outfit. Keep it contemporary, realistic, and suitable for a daytime city walk.

DOG:
Use exactly the requested breed, color, size, and number of dogs. Keep their anatomy, proportions, fur, and scale highly realistic. The dog(s) should naturally walk beside the person.

PHOTOGRAPHY:
Full-body vertical composition, 9:16, eye-level street photography, natural walking motion, warm daylight, soft realistic shadows, subtle depth of field, realistic skin and fabric texture, detailed dog fur, 35mm lens look, premium photorealistic lifestyle photography.

Avoid altered identity, distorted hands, extra fingers, duplicated dogs, incorrect leashes, unrealistic anatomy, crowded background, CGI appearance, artificial skin, exaggerated poses, text, logos, and watermarks.

پس مرحله اول:

مرحله دوم: