立即登录
qwen换衣
25
1
0
API
基础模型: Qwen-Edit
二冰
发布时间: 2025-11-28
关注
提示词
Extract the person from Image 1. Preserve 100% of the person’s identity: facial features, hairstyle, body shape, pose, hands, and any objects they are holding. Do NOT add or remove any items. Do NOT change the pose. Do NOT change the hairstyle. Do NOT generate new accessories. From Image 2, extract ONLY the clothing that appears in each image. Apply the clothing to the person exactly as follows: - If an image shows only a top, replace only the top. - If an image shows a full outfit, replace only the visible parts shown. - Do NOT change pants or shoes unless the clothing image clearly includes pants or shoes. - Do NOT invent or auto-complete missing clothing pieces. When fitting the new clothing on the person from Image 1: - Maintain the original person’s body proportions and silhouette. - Do not distort limbs or hands. - Ensure the person is still holding the same items in the same way. Generate a medium-shot portrait of the person from Image 1 wearing the clothes from Image 2. Lighting, style, perspective, and background must remain consistent with Image 1. Final output: A realistic medium-shot portrait of the original person from Image 1, **wearing exactly the clothing shown in Image 2**, with no other changes.
图像
图像
移位
强度
种子
步数
cfg
采样器名称
调度器
降噪
dimensions
invert
Off
batch_size
提示词
blurry, low quality, low resolution, bad anatomy, deformed, distorted, extra limbs, missing limbs, extra fingers, fused fingers, text, watermark, logo, artifacts, overexposed, underexposed, grainy
运行
队列选择
使用指南
qwen换衣
返图区
暂无返图
查看返图
任务列表
