This code is licensed under CC BY-NC-SA 4.0. Commercial usage is not permitted; please contact [email protected] or [email protected] regarding commercial licensing. If you use this dataset or the code in a scientific publication, please cite the following paper:
@inproceedings{FischerECCV2018,
author = {Tobias Fischer and Hyung Jin Chang and Yiannis Demiris},
title = {{RT-GENE: Real-Time Eye Gaze Estimation in Natural Environments}},
booktitle = {European Conference on Computer Vision},
year = {2018},
month = {September},
pages = {339--357}
}
This work was supported in part by the Samsung Global Research Outreach program, and in part by the EU Horizon 2020 Project PAL (643783-RIA).
More information can be found on the Personal Robotic Lab's website: https://www.imperial.ac.uk/personal-robotics/software/.
- pip:
pip install tensorflow-gpu keras numpy scipy<=1.2.1 tqdm matplotlib pyamg - conda:
conda install tensorflow-gpu keras numpy scipy<=1.2.1 tqdm matplotlib pyamg
This code was used to inpaint the region covered by the eyetracking glasses. There are two parts:
- training subject-specific GANs using the images where no eyetracking glasses are worn (
GAN_train.pyandGAN_train_run.ipynb) and - the actual inpainting using the trained GANs (
GlassesCompletion.pyandGlassesCompletion_run.py).
In GAN_train_run.ipynb and GlassesCompletion_run.py the dataset_folder_path needs to be adjusted to where the dataset was downloaded to.
- ./external/poissonblending.py: MIT License; Link to GitHub
- Some code taken from DC-GAN: MIT License; Link to GitHub
- Tensorflow; Apache License 2.0, Link to website
- Keras; MIT License, Link to website