Workflow Name: Cartoon Blind Box Image
【Workflow Introduction】
Using this workflow, first upload your image photo at the loading image location on the bottom right. Then, use the Florence2 node to reverse-engineer the prompt words, apply style transfer through IPAdapter, and achieve facial feature similarity using Apply PuLID. The sampler's built-in parameters are set and ready to run. Give it a try now!

【Use Case】
You can create your own cartoon blind box image, which allows you to express yourself, showcase your unique style and preferences, and also serves as a commemorative piece. It adds color to life, brings creativity and surprises, and is an excellent way to showcase personality and convey emotions.

【Key Nodes】
IPAdapter, Apply PuLID, Florence2

【Model Version】
SDXL
Model Name: Bamboo Feather Original Full-Function Actual Texture ZY SDXL

【LoRA Model】
Model Name: IP Design Cute National Style First Box 1.0.safetensors

【ControlNet Application】
None

【K Sampler】
CFG: 6.5
Sampling Method: dpmpp_sde
Scheduler: karras
Denoising: 1