search

TadasBaltrusaitis / CLM-framework

CLM-framework (a.k.a Cambridge Face Tracker) is a framework for various Constrained Local Model based face tracking and landmark detection algorithms and their extensions/applications. Includes CLM-Z and CLNF.
Other
471 stars 246 forks source link

how to build a new model #35

Open hualan opened 8 years ago

hualan commented 8 years ago

hi, i've got other datasets, but i don't know how to build model, because eyes landmarks detecting works not good when the eyes is close, thk u!

TadasBaltrusaitis commented 8 years ago

Hi,

It depends on what part of the model you want to retrain. Keep in mind you will need a reasonable amount of well labelled data to perform the training.

To train the models follow the instructions here: https://github.com/TadasBaltrusaitis/CLM-framework/tree/master/matlab_version/pdm_generation - code for training the Point Distribution Model (PDM) You can find the patch training code here - https://github.com/TadasBaltrusaitis/CCNF https://github.com/TadasBaltrusaitis/CLM-framework/tree/master/matlab_version/face_validation - A module for validating face detections (training and inference), it is used for tracking in videos so as to know when reinitialisation is needed

Thanks, Tadas