ChatGPT Image 2
Prompt example
Use the uploaded portrait Image 1 as the sole identity reference. Preserve her face shape and hairstyle as accurately as possible, including facial contours, cheek shape, jawline, chin, eye shape, nose, lips, skin tone, and the center-parted double braids. Ensure the real person is clearly recognizable. Create a high-quality vertical realistic street portrait inspired by a real girl imitating the pose of a 2D anime mural behind her. Scene: Outdoors, under daylight; the background is a clean gray concrete wall with sharp paneled lines, an asphalt road at the bottom, and a blue sky with white clouds above. Character: Image 2. The mural on the wall features a large, flat anime-style character in a famous original pose. In the foreground, the real girl mimics the exact pose and expression of the anime mural, including the head position. She wears her original clothing but replicates the pose and expression perfectly. Style: Clean realism for the real girl, flat anime illustration style for the mural character, strong contrast between reality and anime, bright daylight, natural shadows, exquisite social media street-style photography, accurate human anatomy, realistic hands, realistic clothing folds, and simple composition. No watermarks, no AI-generated tags, no platform IDs, no random text. Negative Prompt: Identity mismatch, different face shape, different hairstyle, missing legs, extra legs, incorrect leg posture, uncoordinated leg posture, deformed hands, extra fingers, facial distortion, blurred face, anatomical errors, incorrect clothing, different school uniform, messy background, watermarks, AI tags, random text, platform ID, Logo, duplicate characters, damaged mural, poor quality. 3:4