🔧 Load CLIPSeg Models

Documentation

• Class name: LoadCLIPSegModels+
• Category: essentials/segmentation
• Output node: False

This node is designed to load the CLIPSeg models for image segmentation, specifically initializing and returning the CLIPSeg processor and model pre-trained on a specific dataset. It abstracts the complexity of model loading, providing an easy-to-use interface for obtaining the necessary components for CLIPSeg-based segmentation tasks.

Input types

Required

Output types

• clip_seg
  • Comfy dtype: CLIP_SEG
  • The output is a tuple containing the CLIPSeg processor and model, ready for use in image segmentation tasks.
  • Python dtype: Tuple[CLIPSegProcessor, CLIPSegForImageSegmentation]

Usage tips

• Infra type: GPU
• Common nodes: unknown

Source code

class LoadCLIPSegModels:
    @classmethod
    def INPUT_TYPES(s):
        return {
            "required": {},
        }

    RETURN_TYPES = ("CLIP_SEG",)
    FUNCTION = "execute"
    CATEGORY = "essentials/segmentation"

    def execute(self):
        from transformers import CLIPSegProcessor, CLIPSegForImageSegmentation
        processor = CLIPSegProcessor.from_pretrained("CIDAS/clipseg-rd64-refined")
        model = CLIPSegForImageSegmentation.from_pretrained("CIDAS/clipseg-rd64-refined")

        return ((processor, model),)