/MistNet

MistNet convolution neural network for segmentation of rain and biology in weather radar data

MIT LicenseMIT

MistNet

MistNet is a convolution neural network for segmentation of rain and biology in weather radar data, developed by Tsung-Yu Lin et al. at University of Massachusetts Amherst

This repository contains a PyTorch implementation of MistNet.

For details, see our publication:

Tsung‐Yu Lin, Kevin Winner, Garrett Bernstein, Abhay Mittal, Adriaan M. Dokter, Kyle G. Horton, Cecilia Nilsson, Benjamin M. Van Doren, Andrew Farnsworth. Frank A. La Sorte, Subhransu Maji, Daniel Sheldon (2019) MistNet: Measuring historical bird migration in the US using archived weather radar data and convolutional neural networks, Methods in Ecology and Evolution, DOI 10.1111/2041-210X.13280

To use MistNet in the vol2bird algorithm, a local copy of mistnet_nexrad.pt is required in addition to the full installation of vol2bird, see vol2bird install instructions for details.