This project implements Image Inpaiting using Deep learning with Generative Adversial Networks.
It uses Cifar10 dataset to reconstruct the center part of the image.
All conda
packages described on requirements.txt
- Run
main.py
to train the model. If you have saved models, add--load-model
argument to load them
Maintainers: Marcelo Andrade marceloga1@al.insper.edu.br
Tickets: Can be opened in Github Issues.
This repository was used as a boilerplate to create these models.
Some hyperparameters were changed according to this paper.
- Implement code on Places dataset.
This project is licensed under MIT license - see LICENSE.md for more details.