search

jerryhouuu / Face-Yaw-Roll-Pitch-from-Pose-Estimation-using-OpenCV

This work is used for pose estimation(yaw, pitch and roll) by Face landmarks(left eye, right eye, nose, left mouth, right mouth and chin)
MIT License
266 stars 60 forks source link

About the chin coordinate #10

Closed Joooshua closed 5 years ago

Joooshua commented 5 years ago

Just getting to know about MTCNN face landmark detection. Sorry to bother if it's a rookie question.

I am wondering how are the chin coordinates obtained. Isn't MTCNN can only generate 5 points (both eyes, nose tip and mouth corners) for one face?

Or should I use Dlib to obtain those coordinates?

Joooshua commented 5 years ago

could you please provide the code that is fine-tuned to get the sixth point?

Oktai15 commented 5 years ago

@Joooshua you can use any 68-landmark detector and take from it coordinate for chin (e.g, this repo: https://github.com/1adrianb/face-alignment)

Joooshua commented 5 years ago

@Joooshua you can use any 68-landmark detector and take from it coordinate for chin (e.g, this repo: https://github.com/1adrianb/face-alignment)

I thought there might be some other ways to change the output of the MTCNN. Thanks, though. I will try.