OpenZhenZhen AI Prompt Gallery · gpt-image-2.5 · image

Photo Restoration Preserving Faces and Expressions

Photo Restoration Preserving Faces and Expressions — explore this gpt-image-2.5 creation on OpenZhenZhen, with its original prompt and a reusable generation example.

Photo Restoration Preserving Faces and Expressions — AI image creation
Photo Restoration Preserving Faces and Expressions

Prompt

Chinese prompt

使用完整已配置 ref-10a-v1 图像作为原始照片输入及面部身份参考,修复整张照片。保留主体精确可辨的面部、五官细节、表情、姿势、服装和原背景;不要重塑五官、改变表情或替换场景。仅修复照片表面实际存在的撕裂、折痕、划痕或人为污损。不要消除人物面部自然纹理或细纹,也不要添加原本不存在的损伤。自然、忠实地恢复色彩,提升清晰度与锐度,并采用平衡的高动态范围摄影效果,避免光晕和人工细节。呈现尼康单反照片般的精致质感,但不声称特定拍摄设备,也不保证数值输出分辨率。保持为同一张完整照片,不添加人物、道具、文字、标志或水印。

English prompt

Restore the complete provided photograph using the full configured ref-10a-v1 image as both the original image input and the facial identity reference. Preserve the subject's exact recognizable face, facial details, expression, pose, clothing and original background; do not reshape features, alter expression or replace the scene. Repair only damage actually present on the photograph itself, such as tears, creases, scratches or human-caused surface marks. Do not remove natural lines or texture from the person's face, and do not add damage that is absent. Recover natural, faithful color, improve clarity and sharpness, and apply a balanced high-dynamic-range photographic finish without halos or artificial detail. Give it the polished look of a Nikon DSLR photograph without claiming a specific capture device or guaranteed output resolution. Keep the result as the same intact photograph, with no added people, props, text, logos or watermark.

View API model

Explore more AI image and video prompts · Try image generation · Try video generation · Create on the AI canvas