Skip to content
This repository has been archived by the owner on May 28, 2024. It is now read-only.

Source code for Fathony, Sahu, Willmott, & Kolter, "Multiplicative Filter Networks", ICLR 2021.

License

Notifications You must be signed in to change notification settings

boschresearch/multiplicative-filter-networks

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

4 Commits
 
 
 
 
 
 

Repository files navigation

Multiplicative Filter Networks

This repository contains a PyTorch MFN implementation and code to perform & reproduce experiments from the ICLR 2021 paper Multiplicative Filter Networks by Rizal Fathony, Anit Kumar Sahu, Devin Willmott, and J. Zico Kolter.

Requirements

  • pytorch 1.7.0
  • torchvision 0.8.1
  • numpy 1.18.1
  • pillow 6.2.1
  • scikit-image 0.16.2

Contents

The file mfn/mfn.py contains implementations of our two instantiations of multiplicative filter networks: FourierNet (Section 3.1) and GaborNet (Section 3.2). It also contains an MFN base class into which any filter may be plugged in (see documentation for details).

The experiments directory contains scripts that correspond to experiments from the paper. Currently, this has:

  • the cameraman image representation experiment from Section 4.1 (image_rep.py), and
  • the cat video representation experiment from Section 4.1 (video_rep.py); see the paper supplement for details on the particular video used

Scripts to reproduce more experiments from the paper will be added soon!

License

"Multiplicative Filter Networks" is open-sourced under the AGPL-3.0 license. See the LICENSE file for details.

About

Source code for Fathony, Sahu, Willmott, & Kolter, "Multiplicative Filter Networks", ICLR 2021.

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages