search

610265158 / Peppa_Pig_Face_Landmark

A simple face detect and alignment method, which is easy and stable.
Apache License 2.0
531 stars 116 forks source link

keypoints.tflite outpute details #37

Closed milad-4274 closed 3 years ago

milad-4274 commented 3 years ago

Hi, Thanks for your awesome repository. I compared your keypoints.tflite model to mediapipe face_landmark.tflite provided by mediapipe. mediapipe output_details has 2 row containing landmark points and confidence(not sure).

but keypoints.tflite model output_details has 3 row. I don't know what are the first 2 rows but the third row is landmark points. is the first 2 rows related to confidence? what are they?

610265158 commented 3 years ago

hi, there is no confidence, the others are head pose, and eyes, mouth state;

milad-4274 commented 3 years ago

Thank you, can you help me to understand what is confidence in landmark? or is it possible to add confidence to this model by retraining it? As far as I know, confidence is related to face detection not landmark points.

610265158 commented 3 years ago

Hi，

In some face landmark model, they treat all the points as independent, so there is a confidence for them.

Simple way, you can regard it as the visibility of each point, same meaning with face detection.