raghavian / mltn

Official code for "Multi-layered tensor networks for image classification", Selvan et al. 2020

README

This is the official Pytorch implementation of "Multi-layered tensor networks for image classification", Raghavendra Selvan et al. 2020

What is this repository for?

• Run and reproduce results in the paper on LIDC dataset
• v1.0

How do I get set up?

• Basic Pytorch dependency
• Tested on Pytorch 1.3, Python 3.6
• Download the data from here
• Unzip the data and point the path to --data_path
• How to run tests: python train.py --data_path data_location --mltn --bn

Usage guidelines

• Kindly cite our publication if you use any part of the code

@misc{raghav2020mltn,
  title={Multi-layered tensor networks for image classification},
  author={Raghavendra Selvan, Silas Ørting and Erik B Dam},
  howpublished={First Workshop on Quantum Tensor Networks in Machine Learning. In conjunction with 34th NeurIPS},
  month={Dec},
  url={https://arxiv.org/abs/2011.06982},
  note={arXiv:2011.06982},
  year={2020}}

Who do I talk to?

• [e-mail]

Thanks to the following repositories we base our project on:

• Torch MPS for the amazing MPS in Pytorch implementations
• Prob.U-Net for preprocessing LIDC data
• Dense Net implementation