nipandha/face-segment

This is the implementation of ‘Scaled Neural Style Transfer for Bounded Face Regions’

I) To create a regular Neural Style Image Transfer:

1. Run the following command with these arguments provided: usage: nst_baseline.py [-h] [--height H] [--width W] [--input I] [--style S] [--num_iter N]

Neural Style Transfer

optional arguments: -h, --help show this help message and exit --height H image height --width W image width --input I input content image --style S style image --num_iter N number of iterations to run the style transfer II) To create a masked face style transfer:

1. Generate a mask file for the face image by running the following: usage: face_detection.py [-h] [--fname F]

Face Detection

optional arguments: -h, --help show this help message and exit --fname F image to be masked,output stored at name_mask.jpg

2. Run the Neural Style Transfer giving the mask file as input mask_img: usage: neural_style_transfer.py [-h] [--height H] [--width W] [--factor FC] [--mask_img M] [--input I] [--style S] [--num_iter N]

PyTorch Gentle Neural Style Transfer for Faces

optional arguments: -h, --help show this help message and exit --height H image height --width W image width --factor FC factor of making style less pronounced on mask --mask_img M mask image --input I input content image --style S style image --num_iter N number of iterations to run the style transfer

Contact: [email protected], [email protected]

face-segment