Use the uploaded image as the ONLY visual reference. Edit this exact photograph rather than recreating the entire scene from scratch.
Preserve the original camera angle, full-body framing, walking pose, hand placement, runway, green hedge, photographers, spectators, depth of field, daylight, shadows and overall composition as closely as possible.
FACE REVEAL: Replace ONLY the completely black, featureless facial covering with a newly generated, naturally beautiful photorealistic female face. Give her a glamorous Zendaya-inspired high-fashion beauty aesthetic and red-carpet presence, without making the result an exact identity replica. Elegant almond-shaped eyes, refined facial structure, realistic lips, natural skin texture, subtle luminous makeup and a confident, composed expression. The face must fit naturally inside the existing hood with realistic shadows around the forehead and sides of the face. Do not make the face look pasted on, overly smooth or artificial.
FIGURE: Preserve and noticeably emphasize the bold, voluptuous hourglass silhouette of the reference: a clearly defined narrow waist, fuller bust, pronounced rounded hips and fuller feminine curves. The fitted fabric should naturally contour the waist and hips, making the hourglass shape clearly visible even beneath the draping. Keep the proportions anatomically believable and elegant. Do not slim the body, flatten the curves, narrow the hips, make the torso boxy, or change her into a straight silhouette.
OUTFIT: Keep the exact same gown design and construction from the reference — same fitted long-sleeve bodice, hood, long cape, ruched waist and hips, asymmetric draping, floor-length skirt, folds and flowing train. Do NOT invent a new dress or add extra fabric. Recolour the existing black fabric into a luxurious deep emerald green, maintaining realistic highlights, shadows, folds and fabric texture. Keep the long glossy gloves black, exactly as in the reference.
Keep the original walking stance and full dress visible from hood to floor. Preserve the dramatic couture feeling of the original photograph.
The final image should look like a real edited version of the original event photograph, as though only the facial covering was removed and the existing gown was recoloured. Authentic paparazzi/editorial photography, realistic skin pores, natural facial detail, realistic fabric physics, subtle photographic grain, natural daylight and shallow depth of field.
Do not: redesign the gown, change the pose, reduce the curves, alter the body into a slim/straight figure, change the background, remove the hood or cape, change the gloves, add jewellery, add exposed skin to the outfit, change the camera angle, crop the dress, create excessive hair covering the outfit, use plastic-looking skin, fantasy styling, cartoon features, or an obviously generated appearance.

PhotoGPTのAI画像

Change outfit

プロンプト

Use the uploaded image as the ONLY visual reference. Edit this exact photograph rather than recreating the entire scene from scratch. Preserve the original camera angle, full-body framing, walking pose, hand placement, runway, green hedge, photographers, spectators, depth of field, daylight, shadows and overall composition as closely as possible. FACE REVEAL: Replace ONLY the completely black, featureless facial…

サイズ1216px x 1536px

参照画像

創作を始める

数千人のアクティブユーザーが集まるDiscordコミュニティに参加して、最新のプロンプトアイデアやコミュニティ作品をチェックしましょう

画像を生成 Discordに参加

Google Veo 3.1 Seedance 2