Mapping of (face detection) bounding box points, from dlib output to final points used/stored in bbs folder.

trigeorgis / mdm

A TensorFlow implementation of the Mnemonic Descent Method.

BSD 3-Clause "New" or "Revised" License

124 stars 48 forks source link

Mapping of (face detection) bounding box points, from dlib output to final points used/stored in bbs folder. #8

Closed gp1313 closed 7 years ago

gp1313 commented 7 years ago

@trigeorgis Can you provide the mapping performed on dlib output (for frontal face detection) to get bounding box used for facial key point prediction.

superphil0 commented 7 years ago

This model, as far as I know was trained on rectangular hull of ground truth face points. If you want to use dlib bboxes make sure they are tight and retain the model ( works best I have tried many variants)