RC1 - detect landmarks script

[endersaka]

Hi Deep3DFaceRecon_pytorch creator and contributors,

In this branch I added a script detect_landmarks.py to batch detect landmarks of custom images using MTCNN. For further informations please read the messages added to RC1 branch and the branches that I merged in it.

Thanks

[sicxu]

Thanks for you contribution! Before I merge your code to the branch, I have some suggestions about the script.

1. Could you add a flag to handle the debugging section? Since it would be unnecessary to draw landmarks for every input image.

2. Please avoid using cv2.imshow as it is inconvenient for most Linux server users. You can change to save the visualization image instead.

3. Could you follow our "data_preparation.py" to set the input arguments? Namely:

python data_preparation.py --img_folder <folder_to_training_images>
python data_preparation.py --img_folder <folder_to_training_images1> <folder_to_training_images2>
python data_preparation.py --img_folder <folder_to_validation_images> --mode=val\

4. You may also modify the data preparation section in README file if you want.

[endersaka]

Hello,

Yes I agree with your observations. I will check and make this changes as soon as I can.