Bannière Urbaine avec Personnage Cohérent

A cinematic wide banner demonstrating consistent AI character generation across three distinct visual zones in a single continuous composition. The overall setting is a moody old city street at dusk with tall buildings, wet pavement, parked cars, and a subtle vignette. On the left, a dim, empty city street with cool blue-gray shadows and no people, slightly desaturated and softly blurred. In the center, a sharp square inset framed by a thin white border containing a realistic portrait of [an unnamed stylish woman in her early 30s] with shoulder-length wavy dark brown hair, medium-light skin, natural makeup, and a calm, confident expression, wearing a dark charcoal overcoat and gray knit sweater. On the right, the exact same woman appears full-body in an expanded matching street scene, holding a takeaway coffee cup, wearing the charcoal overcoat, black high-waisted trousers, and a black shoulder bag, illuminated by warm evening streetlights. Photorealistic editorial photography, shallow depth of field, wet cobblestone reflections, cinematic color grading. No text, logos, UI controls, or watermarks.