Skip to content

Latest commit

 

History

History
28 lines (23 loc) · 942 Bytes

README.md

File metadata and controls

28 lines (23 loc) · 942 Bytes

FacialLandmark_MXNet

Facial landmarks training by MXNet. Use for face alignment and so on.

0. Other DL framework

  • The tensorlfow implementation to see here
  • The caffe implementation to see here

1. Environment

  • MXNet: v1.1.0

2. Training and Testing

  • Check common/landmark_helper.py to read yourself dataset.
  • Check training/training.py. Use python training.py --help to see help.
  • The test script in testing folder.

3. Result image

original original original original original original original original original original original