Giter Club home page Giter Club logo

mcflow's Introduction

MCFlow

A PyTorch implementation of our CVPR 2020 conference paper "MCFlow: Monte Carlo Flow Models for Data Imputation."

By Trevor W. Richardson, Wencheng Wu, Lei Lin, Beilei Xu and Edgar A. Bernal.

[Rochester Data Science Consortium]

[arXiv Paper]

Install and Run Code

  1. Setup repo
  git clone https://github.com/trevor-richardson/MCFlow
  cd MCFlow
  1. Setup environment
  conda env create -f env.yml
  conda activate mcflow
  1. Run experiment
  python main.py

Citation

Please cite the following if you find our work useful in your research:

@InProceedings{Rich_2020_CVPR,
author = {Richardson, Trevor W. and Wu, Wencheng and Lin, Lei and Xu, Beilei and Bernal, Edgar A.},
title = {MCFlow: Monte Carlo Flow Models for Data Imputation},
booktitle = {The IEEE Conference on Computer Vision and Pattern Recognition (CVPR)},
month = {June},
year = {2020}
}

mcflow's People

Contributors

trevor-richardson avatar

Stargazers

 avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar

Watchers

 avatar  avatar  avatar  avatar

mcflow's Issues

benchmark on CIFAR-10 dataset

Hi,

Thanks for your work! I wanted to reproduce the results of your paper on CIFAR-10 dataset, so I added the following code to the path_to_matrix function:

def path_to_matrix(path):
    ## other codes omitted
    elif path == 'cifar':
        trainset = torchvision.datasets.CIFAR10(root='./data', train=True, download=True)
        cifar_train = []
        for idx in range(len(trainset)):
            cifar_train.append(np.moveaxis(np.array(trainset[idx][0]), 2, 0).flatten() / 255)
        
        testset = torchvision.datasets.CIFAR10(root='./data', train=False, download=True)
        cifar_test = []
        for idx in range(len(testset)):
            cifar_test.append(np.moveaxis(np.array(testset[idx][0]), 2, 0).flatten() / 255)
        return cifar_train, cifar_test, (3, 32, 32)
    ## other codes omitted

However, the RMSE is quite different from what was reported. For example, RMSE is about 0.11 when the missing rate is 0.1. So I guess that some preprocessing steps are needed to reproduce the results. Could you please share those steps or the codes to run the CIFAR experiments?

Thanks,

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    ๐Ÿ–– Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. ๐Ÿ“Š๐Ÿ“ˆ๐ŸŽ‰

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google โค๏ธ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.