Consistent Character Maker
created a year ago
character
img2img
controlnet
face
ipadapter
upscale
lora
56 nodes

3.9k

774

Outputs
example image output 0 for workflow.
example image output 1 for workflow.
example image output 2 for workflow.
example image output 3 for workflow.
example image output 4 for workflow.
example image output 5 for workflow.
example image output 6 for workflow.
example image output 7 for workflow.
example image output 8 for workflow.
example image output 9 for workflow.
example image output 10 for workflow.
example image output 11 for workflow.
example image output 12 for workflow.
example image output optimized_comfy_workflows_user_uploads/21088d3e-15a8-49b6-a5a5-a1cebd92eb34/assets/Picture 4.webp for workflow.
Description

What is the biggest problem with AI? consistency... well I made a workflow that remedies this problem. Prompts are not required, but can help.

the principle is simple, you choose a photo of the face, of (clothing / body), a Pose (the mask will be taken on this pose, so choose a model that matches the shape of your model) and a background photo.

Q and enjoys :)

The result is not always perfect, you can play with the denoise and all.

But in principle, the result is almost perfect.

Built-in nodes
Custom nodes
ControlNetLoader
ControlNetApply
PreviewImage
OpenposePreprocessor
RegionalIPAdapterColorMask //Inspire
UpscaleModelLoader
CheckpointLoaderSimple
VAELoader
CLIPSetLastLayer
LoraLoaderModelOnly
ToIPAdapterPipe //Inspire
ApplyRegionalIPAdapters //Inspire
EmptyLatentImage
ImpactImageInfo
ImageScaleToTotalPixels
ImpactKSamplerBasicPipe
ToBasicPipe
FromBasicPipe
VAEDecode
LoadImage
UltimateSDUpscale
SaveImage
MaskComposite
ToBinaryMask
MaskToImage
CLIPTextEncode
SolidMask
ImageToMask
OneFormer-COCO-SemSegPreprocessor
CLIPVisionLoader
IPAdapterModelLoader
SAMLoader
ToDetailerPipe
Models
1
0
3
0