Skip to content
"Learning Discrete and Continuous Factors of Data via Alternating Disentanglement" accepted at ICML2019
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
Dsprites_exp initial commit May 6, 2019
config
tfops
utils initial commit May 6, 2019
.gitignore
LICENSE
README.md Update README.md May 25, 2019
ex.yml initial commit May 6, 2019

README.md

Learning Discrete and Continuous Factors of Data via Alternating Disentanglement

Dependency

  • python=3.5
  • tensorflow version = 1.4
  • CUDA 8.0
  • cuDNN 6.0
  • Environment detail is listed in `ex.yml'

Citing this work

@inproceedings{jeongICML19,
    title={
        Learning Discrete and Continuous Factors of Data via Alternating Disentanglement
    },
    author= {Yeonwoo Jeong and Hyun Oh Song},
    booktitle={International Conference on Machine Learning (ICML)},
    year={2019}
}

Dataset(dSprites)

Edit path

  • Edit path in 'config/path.py'
  • ROOT - (directory for experiment result)
  • DSPRITESPATH - (directory for downloaed dsprites)

Run model

  • Dsprites_exp/CascadeVAE/main.py
  • Dsprites_exp/CascadeVAE-C/main.py

License

MIT License

You can’t perform that action at this time.