The code for Dynamic Orientaion Determination, which is the novelty of the paper, is in the README - only need to add the initial setup when load the dataset and enforce DOD using classify_direction function.
The results can be reproduced using HRNet (https://github.com/HRNet/HRNet-Facial-Landmark-Detection)
I can help you to integrate the code (or, if needed , upload the version over HRNet - just need to validate if it ok with the license)
You can send me mail and I will try to help you there : netana03@cs.huji.ac.il
The code for Dynamic Orientaion Determination, which is the novelty of the paper, is in the README - only need to add the initial setup when load the dataset and enforce DOD using classify_direction function. The results can be reproduced using HRNet (https://github.com/HRNet/HRNet-Facial-Landmark-Detection) I can help you to integrate the code (or, if needed , upload the version over HRNet - just need to validate if it ok with the license) You can send me mail and I will try to help you there : netana03@cs.huji.ac.il