EBP: Energy-Based Processes for Exchangeable Data

Code for "Energy-Based Processes for Exchangeable Data" by Mengjiao Yang, Bo Dai, Hanjun Dai, Dale Schuurmans.

Paper available at https://arxiv.org/abs/2003.07521

Upon using this codebase, please cite the paper:

@article{yang2020energy,
  title={Energy-Based Processes for Exchangeable Data},
  author={Yang, Mengjiao and Dai, Bo and Dai, Hanjun and Schuurmans, Dale},
  journal={arXiv preprint arXiv:2003.07521},
  year={2020}
}

Examples

Multi-modal synthetic data

Visualization of the learned energy of Gaussian Processes (GPs), Nueral Processes (NPs), Variational Implicit Processes (VIPs), and energy-based processes (EBPs). EBPs successfully capture multi-modality of the toy data.

Image completion

Samples of generated MNIST digits conditioned on a subset of pixels:

Point-cloud generation

Point-cloud generation using the learned RNN sampler:

Point-cloud denoising

Point-cloud denoising using the learned energy function:

Install

Navigate to the root of project, and perform:

pip3 install -e .

The installation requires both gcc and CUDA (if gpu is enabled)

Training

cd ebp/experiments/
./run_ebp.sh

Test

To plot the energy heatmap, pass the latest check-pointed epoch number

./run_ebp.sh
./run_ebp.sh -epoch_load 99