pretrained models for 3D semantic image segmentation with nnU-Net
Fabian Isensee, Paul F. Jäger, Simon Köhl, Jens Petersen, Klaus Hermann Maier-Hein · Zenodo (CERN European Organization for Nuclear Research) · 2020
nnU-Net is an out-of-the-box segmentation tool for 3D biomedical image data. It automatically configures U-Net based segmentation pipelines that achieve state of the art result across a broad range of datasets. We have evaluated nnU-Net on 10 biomedical segmentation challenges spanning 19 diverse datasets. This repository contains the trained network parameters for these datasets. Instructions for how to use them are provided at: https://github.com/MIC-DKFZ/nnUNet