LatentCLR: A Contrastive Learning Approach for Unsupervised Discovery of Interpretable Directions

Abstract

Recent research has shown that it is possible to find interpretable directions in the latent spaces of pre-trained Generative Adversarial Networks (GANs). These directions enable controllable image generation and support a wide range of semantic editing operations, such as zoom or rotation. The discovery of such directions is often done in a supervised or semi-supervised manner and requires manual annotations which limits their use in practice. In comparison, unsupervised discovery allows finding subtle directions that are difficult to detect a priori. In this work, we propose a contrastive learning-based approach to discover semantic directions in the latent space of pre-trained GANs in a self-supervised manner. Our approach finds semantically meaningful dimensions compatible with state-of-the-art methods.

Cite

Text

Yüksel et al. "LatentCLR: A Contrastive Learning Approach for Unsupervised Discovery of Interpretable Directions." International Conference on Computer Vision, 2021. doi:10.1109/ICCV48922.2021.01400

Markdown

[Yüksel et al. "LatentCLR: A Contrastive Learning Approach for Unsupervised Discovery of Interpretable Directions." International Conference on Computer Vision, 2021.](https://mlanthology.org/iccv/2021/yuksel2021iccv-latentclr/) doi:10.1109/ICCV48922.2021.01400

BibTeX

@inproceedings{yuksel2021iccv-latentclr,
  title     = {{LatentCLR: A Contrastive Learning Approach for Unsupervised Discovery of Interpretable Directions}},
  author    = {Yüksel, Oğuz Kaan and Simsar, Enis and Er, Ezgi Gülperi and Yanardag, Pinar},
  booktitle = {International Conference on Computer Vision},
  year      = {2021},
  pages     = {14263-14272},
  doi       = {10.1109/ICCV48922.2021.01400},
  url       = {https://mlanthology.org/iccv/2021/yuksel2021iccv-latentclr/}
}