Step 1: Upload an image (it can be a portrait, scene, design material, etc.)

Step 2: Adjust the redraw intensity

Step 3: Run with one click, it will automatically reverse-engineer related prompt words from the image and generate similar images to avoid copyright disputes

This workflow is based on flux reverse-engineering prompt words, image-to-image processing, and image cleaning. It incorporates nunchaku acceleration to improve image generation speed and adds a cfg zero node to enhance image quality