Skip to content

Visually predicting aerial flood imagery using neural style transfer

Notifications You must be signed in to change notification settings

jdiaz4302/flood_style_transfer

 
 

Repository files navigation

Flood style transfer

A project by Jeremy Diaz (https://github.com/jdiaz4302) and Kristin Robinson (https://github.com/CUKrob)

Final project for INFO 4604/5604 - Applied Machine Learning

Background

Artistic style transfer is a recent method for applying artistic styles to everyday pictures, such as making a photograph look like it was painted by Vincent van Gogh (Figure 1C). A foundational paper in this field and the approach which this project is based off is "A Neural Algorithm of Artistic Style" by Leon A. Gatys, Alexander S. Ecker, and Matthias Bethge.

Example of artistic style transfer

Our goal

The goal of this project is to see if natural disasters (such as fires and floods) can be effectively used as artistic styles. That is, can we visually predict what a natural disaster may look like?

Methods

We gathered before and after images for 28 flooded locations. One image was used as a "style" image (representing a typical flood) (Figure 2) to apply to other location's before images. We generated output images which serve as compromises between before-image content and style-image style, and then compared these output images (via the same unweighted loss functions) to the true after images.

Flood style image

Artistic style transfer utilizes a pretrained convolutional neural network, therefore there is no "training"; parameters are not tuned, only hyperparameters (such as loss function tradeoff and neural style depth) are. We found the hyperparameter combinations which minimize the loss/error between output images and 20 post-flood images, then used this combination to evaluate performance on 7 test set images. The model-generated outputs of which are below.

Evaluation

Overall the results are pretty impressive! Notably, however, the model will not "add water". Our model primarily gives the scenes a muddy and eroded appearance, suggesting that we are more reasonably predicting the aftermath (rather than the presence) of a flood. Due to this, the predictions are less impression on locations which lack abundant plant life.

Acknowledgements

This project was competed using the PyTorch (http://pytorch.org/) deep learning framework and was inspired by the above-mentioned arxiv paper. Our neural style tranfer code was immensely helped by the following documentation: http://pytorch.org/tutorials/advanced/neural_style_tutorial.html

About

Visually predicting aerial flood imagery using neural style transfer

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Jupyter Notebook 100.0%