search

jhb86253817 / PIPNet

Efficient facial landmark detector
MIT License
412 stars 82 forks source link

How to train on my own dataset? #12

Closed sky-fly97 closed 3 years ago

sky-fly97 commented 3 years ago

Hello， I have a batch of expanded images and corresponding ldmk annotations. I want the model to be trained both on 300W and my own data at the same time. How should I do?

jhb86253817 commented 3 years ago

Hi, could you give more details, e.g., what do you mean by expanded images? Larger than 256x256?

sky-fly97 commented 3 years ago

Hi, could you give more details, e.g., what do you mean by expanded images? Larger than 256x256?

Hi， thanks for your qucik reply， I'm going to use a batch of 256*256 face images generated by GAN to expand the 300W training set for training. I read the preprocess.py carefully. In fact, this should be a simple thing. Only need to add the expanded data in 300W and change a little code. I'm doing it~