InstantID Face Analysis

Documentation

• Class name: InstantIDFaceAnalysis
• Category: InstantID
• Output node: False

The InstantIDFaceAnalysis node is designed to load and prepare a face analysis model for further processing. It focuses on initializing the model with the specified provider (CPU, CUDA, ROCM) and setting it up for face detection and analysis tasks.

Input types

Required

• provider
  • Specifies the computation provider for the face analysis model, affecting where the model's computations are executed (e.g., on CPU, CUDA, or ROCM).
  • Comfy dtype: COMBO[STRING]
  • Python dtype: List[str]

Output types

• faceanalysis
  • Comfy dtype: FACEANALYSIS
  • Represents the initialized and prepared face analysis model ready for performing face detection and analysis.
  • Python dtype: FaceAnalysis

Usage tips

• Infra type: GPU
• Common nodes: unknown

Source code

class InstantIDFaceAnalysis:
    @classmethod
    def INPUT_TYPES(s):
        return {
            "required": {
                "provider": (["CPU", "CUDA", "ROCM"], ),
            },
        }

    RETURN_TYPES = ("FACEANALYSIS",)
    FUNCTION = "load_insight_face"
    CATEGORY = "InstantID"

    def load_insight_face(self, provider):
        model = FaceAnalysis(name="antelopev2", root=INSIGHTFACE_DIR, providers=[provider + 'ExecutionProvider',]) # alternative to buffalo_l
        model.prepare(ctx_id=0, det_size=(640, 640))

        return (model,)