Pixel Shuffle

Contains all the work done on the Pixel Shuffle layer.

Link to PR

#2563

Link to Blogs

• Blog post 1

• Blog post 2

  

  Probably the most intuitive illustration of Pixel Shuffle on the internet.
  I had a lot of fun understanding and making this.

What’s in this folder

1. Pixel_Shuffle_Demo_1.ipynb - understanding the PyTorch implementation of the layer.
2. Pixel_Shuffle_Demo_2.ipynb - comparison between PyTorch/numpy and mlpack/armadillo implementations of the layer.
3. Pixel_Shuffle_Visual_Experiments.ipynb - to recreate the visuals in blog post 2.
4. test.cpp - Armadillo implementation of the layer’s functionality.

How to run

To run the .cpp file, you need to have armadillo installed in your system.

g++ test.cpp -o test -larmadillo && ./test

To run Pixel_Shuffle_Demo_1.ipynb, you would need pytorch and numpy python libraries.

To run Pixel_Shuffle_Demo_2.ipynb, you would need pytorch, numpy and armadillo. To avoid issues with linking the libraries correctly, it is better to run the notebook in Google Colab.

To run Pixel_Shuffle_Visual_Experiments.ipynb, you would need pytorch, matplotlib, seaborn, numpy python libraries.

Google Colab links for notebooks

• Pixel Shuffle Demo 1
• Pixel Shuffle Demo 2
• Pixel Shuffle Visual Experiments