Workflow Name: Dummy/Mannequin to Real Person
【Workflow Introduction】
Open the workflow, and you first need to upload three images: one facial image, one clothing mannequin image, and one background image. These three images will be processed in reverse. First, the SDXL model combined with brushnet and controlnet is used to generate the image for the first time. Then, the FLUX model is used to refine the image. Finally, the FLUX model is used again to repair the face, upscale the image, and generate the final image. All parameters are pre-configured—just click run to start. Give it a try!

【Usage Scenarios】
Upload three images to generate your desired real-life clothing model with a background. Convenient, fast, and practical—unleash your creativity!

【Key Nodes】
SDXL

【Model Version】
SDXL
Model Name: 万享XL 超写实摄影V8.4 V8.4.safetensors
FLUX
Model Name: flux1 dev fp8 Kijai.safetensors

【ControlNet】
control lora canny rank256.safetensors
Start Time: 0.1
End Time: 1.0
Strength: 0.7

【LoRA Model】
None

【Custom Sampler】
Scheduler: sgm uniform
Denoising: 0.5
Sampler: euler ancestral
【K Sampler】
Scheduler: normal
Denoising: 0.45
Sampler: euler ancestral