luca96/dlib-minified-models

A set of minified (but still accurate) models for Dlib

Dlib Minified Models

A set of minified (but still accurate) models for the Dlib library.

These models are particularly suitable for mobile application thanks to the reduced size of the models, and the increased inference speed.

List of available models, by category:

• face_landmarks: contains models to detect various sets of facial landmarks.
  • FL68: full face 68 landmarks; more efficient version of the Dlib detector.
  • EE22: 22 landmarks of eyes and eyebrows.
  • NM30: 30 landmarks of nose and mouth.
  • FC17: 17 landmarks of the face contour.
  • WFLW98: 98 landmarks, full face, trained on the WLFW dataset.

Contributing

See the contributions guidelines.

dlib-minified-models's Issues

work support

hi Luca, I would like to know if you could make me a budget for a more accurate trained model of mouth and eyes. You can write to my email.

[email protected]

About training process

Thank for your work. I'm going to try it, but I have some questions before starting the work. The first, how many time does it take to produce a new model? And is it faster if run on a machine that have a strong GPU or set the number of thread higher than 1?

test shape predictor

HI, i followed the tutorial and make my owns predictors, thanks a lot. The issue comes when i try to tested. The Error of the model is so bigger (262!). I try to measure your eyebrown predictor an have 272. I use the labels_ibug_300W_test.xml of the tutorial.
My predictors already work, but with medium quality.
Best regards.

profile face training

I am getting the iBug dataset that includes profile face annotations.
Can I use your training script to train a detector on profile face landmarks?
Here is a link to the 39 points profile landmarks I am referring to: 39 point landmarks

Stability

Yet I am working on project with dlib 68 face landmarks.dat file. But in many testing results seems very unstable. After some research I found dataset file.

Here are some source where its very stable example.

98 face landmarks: https://wywu.github.io/projects/LAB/WFLW.html

Very stable 68 landmarks: https://github.com/facebookresearch/supervision-by-registration

But can you guide me how can I make ".dat" file with them. Or can you make more stable ".dat" file and share with me?

In nose_mouth_30.dat.bz2 numbers representing nose and mouth are not correct.

In nose_mouth_30.dat.bz2 numbers representing nose and mouth are not in sync with the image shown. Because of that not able to extract only mouth coordinates from landmarks