Photorealistic Street-Fashion Portrait with Face-Lock

GPT image 2 Use my uploaded face photo as the only facial identity reference with strict face-lock preservation (95–100% similarity). Keep the exact same face, eyes, eyebrows, nose, lips, jawline, skin tone, ethnicity, age, proportions, and natural skin texture. Create a real DSLR street-fashion photo of a confident young woman (20–25) with subtle Douyin-style makeup, visible pores, natural skin, long black wavy hair, oversized white sleeveless shirt, light-wash baggy jeans, black-and-white plaid flannel tied around the waist, iced coffee in one hand, layered silver necklace, earrings, and black wristwatch. Match the reference pose exactly: standing in the middle of a modern city street, one hand holding coffee, the other touching her hair, head slightly raised, looking away from the camera. Full-body vertical 9:16 composition, low-angle 35mm DSLR shot, urban downtown background with tall buildings and soft blur, natural daylight, cinematic editorial color grading, ultra-photorealistic RAW photography, realistic hair strands, fabric textures, eye reflections, and 8K fashion-campaign quality. Remove all watermarks, logos, and text. Negative prompt: anime, manga, cartoon, CGI, painting, plastic skin, beauty filters, identity blending, fantasy elements, flowers, indoor scenes, wrong pose, wrong outfit, blurry face, bad anatomy, watermark, logo, text. --ar 9:16 --style raw --quality high