Unsupervised Discovery of Disentangled Manifolds in GANs

By Yu-Ding Lu, Hsin-Ying Lee, Hung-Yu Tseng, Ming-Hsuan Yang. 2020

9 citations recorded by the source corpus.

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.

Subfields: Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

1. Qifeng Chen
2. Jue Wang
3. Tengfei Wang
4. Yanbo Fan
5. Yong Zhang
6. Vladlen Koltun
7. Luc Van Gool
8. Ming-Hsuan Yang

Citations in the corpus

• High-Fidelity GAN Inversion for Image Attribute Editing (2021, background): “Some [22,35,47] found distinguishable directions based on mutual information.”