Regional IPAdapter Encoded Mask (Inspire)

Documentation

• Class name: RegionalIPAdapterEncodedMask __Inspire
• Category: InspirePack/Regional
• Output node: False

This node specializes in applying encoded mask-based image processing adjustments within the InspirePack framework, leveraging regional IP adapter techniques to conditionally modify image embeddings based on specified masks and weights.

Input types

Required

• mask
  • The mask input specifies areas of the image to be targeted for conditional embedding adjustments, playing a crucial role in the node's operation by defining regions for focused processing.
  • Comfy dtype: MASK
  • Python dtype: torch.Tensor
• embeds
  • Embeddings that represent the desired adjustments or features to be applied to the specified regions of the image, influencing the final output based on the mask.
  • Comfy dtype: EMBEDS
  • Python dtype: torch.Tensor
• weight
  • A float value that determines the intensity of the embedding adjustments applied to the image, allowing for fine-tuning of the effect strength.
  • Comfy dtype: FLOAT
  • Python dtype: float
• weight_type
  • Specifies the method of applying weights to the embeddings, offering options like original, linear, or channel penalty for diverse adjustment effects.
  • Comfy dtype: COMBO[STRING]
  • Python dtype: str
• start_at
  • Defines the starting point of the effect application in terms of image processing, enabling phased adjustments.
  • Comfy dtype: FLOAT
  • Python dtype: float
• end_at
  • Sets the endpoint for the effect application, allowing for precise control over the extent of the adjustments.
  • Comfy dtype: FLOAT
  • Python dtype: float
• unfold_batch
  • A boolean flag that, when true, processes each item in a batch individually, enhancing flexibility in handling batched inputs.
  • Comfy dtype: BOOLEAN
  • Python dtype: bool

Optional

• neg_embeds
  • Optional negative embeddings that can be used to specify features or adjustments to be avoided in the specified regions, adding an inverse effect capability.
  • Comfy dtype: EMBEDS
  • Python dtype: torch.Tensor

Output types

• regional_ipadapter
  • Comfy dtype: REGIONAL_IPADAPTER
  • Produces a conditioned version of the input based on the encoded mask and specified parameters, reflecting the targeted adjustments.
  • Python dtype: IPAdapterConditioning

Usage tips

• Infra type: GPU
• Common nodes: unknown

Source code

class RegionalIPAdapterEncodedMask:
    @classmethod
    def INPUT_TYPES(s):
        return {
            "required": {
                "mask": ("MASK",),

                "embeds": ("EMBEDS",),
                "weight": ("FLOAT", {"default": 0.7, "min": -1, "max": 3, "step": 0.05}),
                "weight_type": (["original", "linear", "channel penalty"],),
                "start_at": ("FLOAT", {"default": 0.0, "min": 0.0, "max": 1.0, "step": 0.001}),
                "end_at": ("FLOAT", {"default": 1.0, "min": 0.0, "max": 1.0, "step": 0.001}),
                "unfold_batch": ("BOOLEAN", {"default": False}),
            },
            "optional": {
                "neg_embeds": ("EMBEDS",),
            }
        }

    RETURN_TYPES = ("REGIONAL_IPADAPTER", )
    FUNCTION = "doit"

    CATEGORY = "InspirePack/Regional"

    def doit(self, mask, embeds, weight, weight_type, start_at=0.0, end_at=1.0, unfold_batch=False, neg_embeds=None):
        cond = IPAdapterConditioning(mask, weight, weight_type, embeds=embeds, start_at=start_at, end_at=end_at, unfold_batch=unfold_batch, neg_embeds=neg_embeds)
        return (cond, )