Simple Comfyui Workflow with controlnets — v2.0

This workflow is for converting a toon image into a real-style image. It uses image-to-image processing with Union XL ControlNet, takes an image and prompts, and outputs an image.

V1 is a simple image-to-image flow that uses ControlNet.

V2 adds a face detailer and a LoRA, with the description noting character LoRA use on the output.

The graph names LoadImage, PreviewImage, and SaveImage nodes for image loading, preview, and saving.

The source lists this as a Workflows item by Scorpiov2 and names SDXL 1.0 as its base model.

The listed model files are bbox/face_yolov8m.pt, centerfold_v90Naughty.safetensors, controlnet++_union_sdxl.safetensors, detailer_v5-SDXL.safetensors, and epicrealismXL_vxFinalkiss_969698.safetensors.

Named nodes include AV_ControlNetPreprocessor, AV_LoraLoader, CheckpointLoaderSimple, CLIPTextEncode, ControlNetApply, ControlNetLoader, CR Image Input Switch (4 way), and EmptyLatentImage.

Other named nodes are FaceDetailer, KSampler, UltralyticsDetectorProvider, VAEDecode, LoadImage, PreviewImage, and SaveImage.

Add prompts to steer the result toward what you want to see.

Sugestão · não verificado

Choose V1 for the simple ControlNet flow; choose V2 if you want to apply a character LoRA, since it adds a face detailer and a LoRA.

Sugestão · não verificado

Check that your setup includes the five listed model files before using the workflow.

Sugestão · não verificado

No license is specified for this workflow.