Workflow Introduction

This workflow is deeply optimized based on Stable Diffusion and ComfyUI, focusing on clothing replacement and virtual try-on. It supports various scenarios such as e-commerce product images, design drafts, and real-life models. Through precise masking segmentation and AI image generation technology, it can achieve:

  • One-click outfit change: Retain model poses and backgrounds, seamlessly replace clothing

  • Intelligent cutout: Automatically recognize clothing parts, support manual fine-tuning

  • High-definition output: Support 4K resolution, suitable for printing and online display

  • Consistent style: Built-in style models ensure new clothing naturally blends with scene lighting and materials


Core Features Highlights

✅ Clothing Selector: 20 preset clothing categories (tops/dresses/shoes and bags, etc.), accurately locate replacement areas
✅ Mask Expansion Technology: Intelligently repair blurred edges to avoid harsh seams
✅ Multi-size Adaptation: Automatically match the proportions of models and clothing images to reduce distortion


Three-Step Simplified Operation Guide

Step 1: Upload Clothing Image

  1. Drag in clothing image: Drag a flat clothing image or model display image into the "Clothing Upload Area"

  2. Select clothing parts:

    • In the "Clothing Selector," check the target part (e.g., upper_clothes = top)

    • Support multi-selection for complex designs (e.g., dress belt = dress belt)

Step 2: Edit Model Mask

  1. Upload model image: Drag the model image to be dressed into the "Model Upload Area"

  2. Right-click to enter Mask Editor:

    • Use the brush to paint over the clothing area to be replaced (recommended to extend the edges by 10-20 pixels)

    • Shortcuts: [ ] adjust brush size, Ctrl Z undo mistakes

  3. Optimize mask:

    • Check the "Mask Expansion" parameter (default value 10) to eliminate residual edges

    • Enable "Double Masking" for background protection in complex scenes

Step 3: Generate and Optimize

  1. Click Generate: Wait for AI to automatically synthesize (about 1 minute)

  2. View and Adjust:

    • Check clothing fit and lighting effects in the "Preview Window"

  3. Save Results:


Notes

⚠️ Image Size Recommendations:

  • Ensure resolution of clothing and model images are similar (recommended 1600x1600px)

  • Crop and align smartphone photos first to avoid disproportion


Accelerate creativity with technology, and present every design perfectly! 🚀