Skip to content

Latest commit

 

History

History
9 lines (7 loc) · 635 Bytes

README.md

File metadata and controls

9 lines (7 loc) · 635 Bytes

Image inpainting with GANs

Instructions

  1. To install all dependencies, run pip install pytorch && pip install Pillow && pip install torchvision
  2. To download the dataset, run the DownloadData.py file. This will download the MIT Places Dataset using torchvision's dataset loader.
  3. To create masked images, run the MakeData.py file. This will preprocess the images by adding a mask at the center of the screen.
  4. To run the baseline analytical inpainting solutions, follow the instructions on the AnalyticalSolns.ipynb file.
  5. To run the GAN inpainting solution, follow the instructions on the inpaint.ipynb file.