基于语义分割的图片前景和背景分离ComfyUI工作流 | AI绘画教程(VIP资源)

节点与模型

节点

comfyui-inpaint-nodes节点:https://github.com/Acly/comfyui-inpaint-nodes

模型

SAM模型

sam_hq_vit_h.pth:https://huggingface.co/lkeab/hq-sam/tree/main

存储文件夹:“\ComfyUI_windows_portable\ComfyUI\models\sams”

G-Dino模型

groundingdino_swinb_cogcoor.pth:https://huggingface.co/ShilongLiu/GroundingDINO/tree/main

存储文件夹:“\ComfyUI_windows_portable\ComfyUI\models\grounding-dino”

inpaint模型

big-lama.pt:https://github.com/Sanster/models/releases/download/add_big_lama/big-lama.pt

fooocus_inpaint_head.pth:https://huggingface.co/lllyasviel/fooocus_inpaint/resolve/main/fooocus_inpaint_head.pth?download=true

inpaint_v26.fooocus.patch:https://huggingface.co/lllyasviel/fooocus_inpaint/resolve/main/inpaint_v26.fooocus.patch?download=true

上述3个模型文件存储文件夹:“\ComfyUI_windows_portable\ComfyUI\models\inpaint”

checkpoint大模型

realismEngineSDXL_v20VAE.safetensors:https://civitai.com/models/152525?modelVersionId=258380

WD14反推提示词模型

下载的模型一定要重命名为下面的名称

wd-v1-4-moat-tagger-v2.onnx:https://huggingface.co/SmilingWolf/wd-v1-4-moat-tagger-v2/resolve/main/model.onnx?download=true

wd-v1-4-moat-tagger-v2.csv:https://huggingface.co/SmilingWolf/wd-v1-4-moat-tagger-v2/resolve/main/selected_tags.csv?download=true

上述2个文件的存储文件夹“\ComfyUI_windows_portable\ComfyUI\custom_nodes\ComfyUI-WD14-Tagger\models”

IPAdapter模型

ip-adapter-plus_sdxl_vit-h.safetensors:https://huggingface.co/h94/IP-Adapter/resolve/main/sdxl_models/ip-adapter-plus_sdxl_vit-h.safetensors?download=true

存储文件夹:“\ComfyUI_windows_portable\ComfyUI\models\ipadapter\SDXL”

CLIP_VISION模型

pytorch_model.bin:https://huggingface.co/h94/IP-Adapter/resolve/main/models/image_encoder/pytorch_model.bin?download=true

存储文件夹:“\ComfyUI_windows_portable\ComfyUI\models\clip_vision\SD1.5\image_encoder”

需要此工作流可以点击下面下载↓↓↓

资源下载此资源下载价格为25草币,请先

原创文章,作者:朋远方,如若转载,请注明出处:https://caovan.com/jiyuyuyifengedetupianqianjinghebeijingfenlicomfyuigongzuoliu-aihuihuajiaocheng/.html

Like (0)
Donate 微信扫一扫 微信扫一扫
朋远方的头像朋远方
Previous 2024年2月5日 下午3:17
Next 2024年2月6日 下午10:51

相关推荐

发表回复

Please Login to Comment