Face Analysis Models

Documentation

Class name: FaceAnalysisModels
Category: FaceAnalysis
Output node: False

The FaceAnalysisModels node is designed to load and configure face analysis models based on specified libraries and providers. It abstracts the complexity of initializing face detection and recognition models, offering a streamlined way to prepare these models for further face analysis tasks.

Input types

Required

library
- Specifies the library to use for face analysis, influencing which face detection and recognition models are loaded and configured.
- Comfy dtype: COMBO[STRING]
- Python dtype: str
provider
- Determines the computational backend (e.g., CPU, CUDA) for the face analysis models, affecting performance and compatibility.
- Comfy dtype: COMBO[STRING]
- Python dtype: str

Output types

analysis_models
- Comfy dtype: ANALYSIS_MODELS
- Outputs a configured face analysis model, ready for performing tasks such as face detection, landmark detection, and face recognition.
- Python dtype: dict

Usage tips

Infra type: CPU
Common nodes: unknown

Source code

class FaceAnalysisModels:
    @classmethod
    def INPUT_TYPES(s):
        return {"required": {
            "library": (INSTALLED_LIBRARIES, ),
            "provider": (["CPU", "CUDA", "DirectML", "OpenVINO", "ROCM", "CoreML"], ),
        }}

    RETURN_TYPES = ("ANALYSIS_MODELS", )
    FUNCTION = "load_models"
    CATEGORY = "FaceAnalysis"

    def load_models(self, library, provider):
        out = {}

        if library == "insightface":
            out = {
                "library": library,
                "detector": FaceAnalysis(name="buffalo_l", root=INSIGHTFACE_DIR, providers=[provider + 'ExecutionProvider',])
            }
            out["detector"].prepare(ctx_id=0, det_size=(640, 640))
        else:
            out = {
                "library": library,
                "detector": dlib.get_frontal_face_detector(),
                "shape_predict": dlib.shape_predictor(os.path.join(DLIB_DIR, "shape_predictor_68_face_landmarks.dat")),
                "face_recog": dlib.face_recognition_model_v1(os.path.join(DLIB_DIR, "dlib_face_recognition_resnet_model_v1.dat")),
            }
        

        return (out, )

2.3 KiB Raw Permalink Blame History