工作流介绍
这个comfyui工作流程是在原有的datou旧照片修复XL工作流的基础上进行优化的。它对Ollama节点中的图像描述提示进行了调整，以确保修复后的照片不会保持黑白状态。此外，为了更精准地恢复原始图像，工作流中为线稿添加了controlnet节点，并用facerestore节点替换了原先的faceswap节点，这样做可以避免在多人场景中出现的问题。这些改进使得整个comfyui工作流程更加高效和稳定。
工作流界面预览
作者：多彩AI
模型详情
Checkpoints
juggernautXL_v9Rdphoto2Lightning.safetensors👉
节点详情
内置节点
DF_Image_scale_to_side
DepthAnything_V2
DownloadAndLoadDepthAnythingV2Model
自定义节点
Comfyroll Studio
– CR Text
ComfyUI
– CLIPTextEncode
– PreviewImage
– ControlNetApplyAdvanced
– ControlNetLoader
– VAEDecode
– CheckpointLoaderSimple
– VAEEncode
– SaveImage
– KSampler
– LoadImage
ComfyUI Bringing Old Photos Back to Life
– BOPBTL_ScratchMask
– BOPBTL_LoadRestoreOldPhotosModel
– BOPBTL_LoadScratchMaskModel
– BOPBTL_RestoreOldPhotos
ComfyUI Easy Use
– easy cleanGpuUsed
ComfyUI Nodes for Inference.Core
– DWPreprocessor
ComfyUI Ollama
– OllamaGenerate
– OllamaVision
comfyui-art-venture
– ImageScaleToMegapixels
ComfyUI-AutomaticCFG
– Automatic CFG
ComfyUI-Gemini
– DisplayText_Zho
– ConcatText_Zho
comfyui-image-round
– ComfyUI_Image_Round__ImageRoundAdvanced
– ComfyUI_Image_Round__ImageCropAdvanced
ComfyUI-Logic
– String
ComfyUI’s ControlNet Auxiliary Preprocessors
– AnyLineArtPreprocessor_aux
FizzNodes
– ConcatStringSingle
ReActor Node for ComfyUI
– ReActorRestoreFace
下载文件夹预览
