
Powered By RTX 4090
Say goodbye to toy-level multimodal editing models generating base images, return to the essence of the king of all things migration, still look at Redux fill, enhanced control and style reference along with character consistency have become the issues we urgently need to solve for generating images at the beginning and end frames, so let's use this to solve it.
Aiwood爱屋研究室
2025-05-12 Update
Aiwood爱屋研究室
2025-05-12 Update
Workflow introduction
Say goodbye to toy-level multimodal editing models generating base images, return to the essence of the king of all things migration, still look at Redux fill, enhanced control and style reference along with character consistency have become the issues we urgently need to solve for generating images at the beginning and end frames, so let's use this to solve it.
Nodes Information
51
CLIPTextEncode
CLIPVisionEncode
CLIPVisionLoader
CheckpointLoaderSimple
ConditioningZeroOut
ControlNetLoader
DualCLIPLoader
EmptyImage
ImageInvert
InpaintModelConditioning
KSampler
LoadImage
LoraLoaderModelOnly
PreviewImage
SaveImage
StyleModelApply
StyleModelLoader
UNETLoader
VAEDecode
VAEEncode
VAELoader
AIO_Preprocessor
ApplyInstantID
ControlNetApplySD3
Crop Face
DifferentialDiffusion
FaceAnalysisModels
FaceSegmentation
FluxGuidance
Get Image Size
GrowMaskWithBlur
Hina.PoseEditor3D
Image To Mask
ImageConcanate
ImageResize+
ImageResizeKJv2
ImageScaleToMegapixels
InpaintCrop
InpaintResize
InpaintStitch
InstantIDFaceAnalysis
InstantIDModelLoader
LayerMask: LoadFlorence2Model
LayerMask: SegmentAnythingUltra V2
LayerUtility: Florence2Image2Prompt
Load RetinaFace
MaskPreview+
PlaySound|pysssss
SetUnionControlNetType
easy cleanGpuUsed
easy imageCropFromMask