As recent generative models can generate photo-realistic images, people seek to understand the mechanism behind the generation process. Interpretable generation process is beneficial to various image editing applications. In this work, we propose a framework to discover interpretable directions in the latent space given arbitrary pre-trained generative adversarial networks. We propose to learn the transformation from prior one-hot vectors representing different attributes to the latent space used by pre-trained models. Furthermore, we apply a centroid loss function to improve consistency and smoothness while traversing through different directions. We demonstrate the efficacy of the proposed framework on a wide range of datasets. The discovered direction vectors are shown to be visually corresponding to various distinct attributes and thus enable attribute editing.
@article{arxiv.2011.11842,
title = {Unsupervised Discovery of Disentangled Manifolds in GANs},
author = {Yu-Ding Lu and Hsin-Ying Lee and Hung-Yu Tseng and Ming-Hsuan Yang},
journal= {arXiv preprint arXiv:2011.11842},
year = {2020}
}