Giters

Maclory / Deep-Iterative-Collaboration

Pytorch implementation of Deep Face Super-Resolution with Iterative Collaboration between Attentive Recovery and Landmark Estimation (CVPR 2020)

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Annotation

HITxyer opened this issue · comments

commented

Hi,thank you very much for your sharing the code, I have some confusion about how the .pkl in the annotation file is generated Looking forward to your reply.

commented

Hi, thanks for your interest.

We use OpenFace to generate 68 landmarks for CelebA faces, and use the official 5-landmark annotation to check the result(We only use 68-landmark annotations whose errors are smaller than a threshold). Based on the landmarks, we calculate the bounding boxes that can properly cover the face. Finally we put the bounding boxes and 68 landmarks in the pkl file.