Skip to content

Latest commit

 

History

History
65 lines (43 loc) · 2.18 KB

README.md

File metadata and controls

65 lines (43 loc) · 2.18 KB

Learning Data Augmentation

OpenAI Request for Research - https://blog.openai.com/requests-for-research-2/

Requirements

  • Python 3
  • Anaconda (Numpy, Matplotlib, PIL, etc)
  • PyTorch
  • GPU, CUDA

Tested on Ubuntu 16.04 with GeForce GTX 1080 Ti

Papers

Approaches

  1. Autoencoder: Encode --> Peturb/Interpolate --> Decode
  2. Synthetic Dataset Generation with CGAN
  3. Domain Adaption / Image-to-Image translation (CycleGAN)
  4. Augmentation / Preprocessing Net + backprop classification loss

Posts

Autoencoder

VAE

GAN

Code

Autoencoder

VAE

Links