Is media pipe is use with deep face for face recognition for better accuracy

Question

I'm using deep face for recognition but the accuracy is not good so I try to implement the media pipe in which I extract the landmarks so I give this to the deep face for better accuracy. Is there any way to do this?

I extract the feature vector from the media pipe but how do I pass this to the deep face? Is there any possible way to do this?

Is media pipe landmark used for deep face for face recognition for better accuracy?

Answer 1

DeepFace already wraps MediaPipe face detector.

result = DeepFace.verify(
   img1_path,
   img2_path,
   detector_backend = "mediapipe"
)
print(result)

But if you want to reach its maximum power, then I recommend you to use Facenet facial recognition model and retinaface face detector pair.

result = DeepFace.verify(
   img1_path,
   img2_path,
   model_name="Facenet",
   detector_backend = "retinaface"
)
print(result)