Skip to content
Keras Implementation of Painting outside the box
Branch: master
Clone or download
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
checkpoint update outpaint, dataloder, add gitignore Jul 28, 2018
data update outpaint, dataloder, add gitignore Jul 28, 2018
saved_images add gitignore to saved_image Jul 29, 2018
.gitignore add gitignore Jul 29, 2018
README.md
augment_image.py add augment_image Jan 1, 2019
dataloader.py
outpaint.ipynb
prepare_data.py update prepare_data, dataloader Jan 1, 2019
prepare_data.sh update prepare_data, dataloader Jan 1, 2019
requirements.txt

README.md

Keras implementation of Image OutPainting

This is an implementation of Painting Outside the Box: Image Outpainting paper from Standford University. Some changes have been made to work with 256*256 image:

  • Added Identity loss i.e from generated image to the original image
  • Removed patches from training data. (training pipeline)
  • Replaced masking with cropping. (training pipeline)
  • Added convolution layers.

Results

The model was train with 3500 scrapped beach data with agumentation totalling upto 10500 images for 25 epochs. Demo

Recursive painting

Demo

Install Requirements

sudo pip3 install -r requirements.txt

Get Started

  1. Prepare Data:
    # Downloads the beach data and converts to numpy batch data
    # saves the Numpy batch data to 'data/prepared_data/'
    sh prepare_data.sh
  2. Build Model
    • To build Model from scratch you can directly run 'outpaint.ipynb'
      OR
    • You can Download my trained model and move it to 'checkpoint/' and run it.

References

You can’t perform that action at this time.