Consistent Character Maker, Using IP-Adapter (2 Character possible)
created a year ago
character
img2img
controlnet
ipadapter
face
lora
125 nodes
5.1k
954
Outputs





















Description
I made a workflow to make consistent character using IP Adapter
The way it work is easy :
Choose a pose image (Note The final result image size will be base on the pose image size)
Choose a background image : Will fill the background of the character or 2 character
Choose a Face Image for the character
Choose a Body Image for the character
Use Positive prompt for Detail : Ex : You want to change the hair, it good to also describe the person you want to make.. i made the picture you will see using :
Pose : 1 man, long blond hair, silver armor
Neg : Blurry Somme Embedding
Built-in nodes
Custom nodes
ControlNetLoader
ControlNetApply
ImpactImageInfo
EmptyImage
UltralyticsDetectorProvider
ToIPAdapterPipe //Inspire
FromBasicPipe_v2
CLIPVisionLoader
ImpactSEGSOrderedFilter
ImpactDilateMaskInSEGS
SegsToCombinedMask
EmptyLatentImage
ImageCompositeMasked
EditBasicPipe
ApplyRegionalIPAdapters //Inspire
PreviewImage
BboxDetectorSEGS
ImageToMask
IPAdapterModelLoader
RegionalIPAdapterColorMask //Inspire
VAELoader
LoraLoader
ToBasicPipe
ImageScaleToTotalPixels
ImpactKSamplerBasicPipe
FromBasicPipe
VAEDecode
SAMLoader
LoadImage
OpenposePreprocessor
SaveImage
CLIPTextEncode
CheckpointLoaderSimple
FaceDetailer
SegmDetectorSEGS
Models