Skip to content
Code for the paper "Generative Modeling of Infinite Occluded Objects for Compositional Scene Representation"
Python Jupyter Notebook Shell
Branch: master
Clone or download
Latest commit 47f8c8e May 15, 2019
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
data
experiments
src
.gitignore
README.md

README.md

Generative Modeling of Infinite Occluded Objects for Compositional Scene Representation

This is the code repository of the paper "Generative Modeling of Infinite Occluded Objects for Compositional Scene Representation".

Dependencies

  • pytorch == 1.0
  • torchvision == 0.2
  • numpy >= 1.15
  • h5py >= 2.8
  • scipy >= 1.1
  • scikit-learn == 0.19
  • matplotlib >= 2.2

Datasets

Change the current working directory to data and run create_datasets.sh.

cd data
./create_datasets.sh
cd ..

Experiments

Change the current working directory to experiments and run run.sh.

cd experiments
./run.sh
cd ..

Run experiments/evaluate.ipynb to evaluate the trained models.

You can’t perform that action at this time.