snowGAN

The snowGAN was an explorator generative adversarial network (GAN) built to learn from the Rocky Mountain snowpack data. Initially the AI was trained to simply understand features of the data, with the intention of transfering what it learned to a new task that could assist in assessing, if not predicting itself, avalanche risk. Explore applying AI to snowpack data and learning about avalanches with us by training your own AI with the dataset, or leveraging the snowGAN alongside the snowGAN as pre-trained models to apply to other problems!

Explore Fine-tuning GANs