Deeper Image Quality Transfer: Low Memory Neural Networks for 3D Images

An illustration of the low-memory technique used in: Deeper Image Quality Transfer: Training Low-Memory Neural Networks for 3D Images http://arxiv.org/abs/1808.05577 by Stefano B. Blumberg, Ryutaro Tanno, Iasonas Kokkinos, Daniel C. Alexander

Contact: stefano.blumberg.17@ucl.ac.uk

Please contact for questions, or on how to apply the low-memory system to other neural networks (e.g. low-memory systems for DeepMedic, UNet, VNet)

Citation

If you find the code and/or paper useful, please cite us as:

@article{DIQT,
author={Blumberg, S B. and Tanno, R. and Kokkinos, I. and Alexander, D C.},
title={Deeper Image Quality Transfer: Training Low-Memory Neural Networks for 3D Images},
journal={Medical Image Computing and Computer Assisted Intervention (MICCAI) 2018},
}

Contents

DIQT_Example.yml : Environment, use conda to install

DIQTExampleMain.py : Main file

models.py : Contains the neural networks