Invert and perturb GAN images for test-time ensembling
GAN Ensembling
Ensembling with Deep Generative Views.
Lucy Chai, Jun-Yan Zhu, Eli Shechtman, Phillip Isola, Richard Zhang
CVPR 2021
Prerequisites
- Linux
- Python 3
- NVIDIA GPU + CUDA CuDNN
Table of Contents:
- Colab – run a limited demo version without local installation
- Setup – download required resources
- Quickstart – short demonstration code snippet
- Notebooks – jupyter notebooks for visualization
- Pipeline – details on full pipeline
We project an input image into the latent space of a pre-trained GAN and perturb it slightly to obtain modifications of the input image. These alternative views from the GAN are ensembled at test-time, together with the original image, in a downstream classification task.
To synthesize deep generative views, we first align (Aligned Input)