MR - Python Implementation
This repositery provides a python implementation of MR (Masked Regression). MR can efficiently synthesize facial expressions. The demo video for MR can be found here.
1 Punjab University College of Information Technology
(PUCIT), Lahore, Pakistan
2 Department of Computer Science, Information Technology
University (ITU), Lahore, Pakistan
International Journal of Computer Vision (IJCV), 2020
2. Create a folder structure as described here.
- Split images into training and test sets (e.g., 90%/10% for training and test, respectively).
- Crop all images to 128 x 128, where the faces are centered.
To train MR:
$ python main.py --mode train --train_dataset_dir 'dataset/train/' --image_size 128 --total_images 200 --input_ch 1
--receptive_field 3 --lamda 0.4 To test MR:
$ python main.py --mode test --test_dataset_dir 'dataset/test/' --image_size 128 --total_images 20 --input_ch 1
--receptive_field 3 To test MR:
$ python main.py --mode test_inthewild --test_dataset_dir 'dataset/inthewild/' --image_size 128 --total_images 20 --input_ch 1
--receptive_field 3 Facial expression synthesis on sketches and animals
Facial expression synthesis on in the wild images
If this work is useful for your research, please cite our Paper:
@article{khan_mr_ijcv_2019,
author="Khan, Nazar and Akram, Arbish and Mahmood, Arif and Ashraf, Sania and Murtaza, Kashif",
journal="International Journal of Computer Vision",
pages = "1433--1454",
title = "{Masked Linear Regression for Learning Local Receptive Fields for Facial Expression Synthesis}",
volume = "128",
year = "2020"
}