Masks to Tensors

Documentation

• Class name: SaltMasksToImages
• Category: SALT/Masking
• Output node: False

This node is designed to convert a collection of mask images into a tensor representation, facilitating the transition from image processing to tensor operations for further computational tasks.

Input types

Required

• masks
  • The masks input represents a collection of mask images that are to be converted into tensor format. This conversion is crucial for enabling subsequent tensor-based operations and analyses.
  • Comfy dtype: MASK
  • Python dtype: List[torch.Tensor]

Output types

• images
  • Comfy dtype: IMAGE
  • The output is a tensor comprising the converted images from the input masks, ready for further processing or analysis.
  • Python dtype: torch.Tensor

Usage tips

• Infra type: GPU
• Common nodes: unknown

Source code

class SaltMasksToImages:
    @classmethod
    def INPUT_TYPES(cls):
        return {
            "required": {
                "masks": ("MASK",)
            }
        }

    RETURN_TYPES = ("IMAGE",)
    RETURN_NAMES = ("images",)

    CATEGORY = f"{NAME}/Masking"
    FUNCTION = "convert"

    def convert(self, masks):
        images = []
        for mask in masks:
            images.append(pil2tensor(mask2pil(mask)))
        images = torch.cat(images, dim=0)
        return (images, )