Latent Space Conditioning on Generative Adversarial Networks

Ricard Durall, Ricard Durall, Ricard Durall, Kalun Ho, Kalun Ho, Kalun Ho, Franz-Josef Pfreundt, Janis Keuper, Janis Keuper

Abstract

Generative adversarial networks are the state of the art approach towards learned synthetic image generation. Although early successes were mostly unsupervised, bit by bit, this trend has been superseded by approaches based on labelled data. These supervised methods allow a much finer-grained control of the output image, offering more flexibility and stability. Nevertheless, the main drawback of such models is the necessity of annotated data. In this work, we introduce an novel framework that benefits from two popular learning techniques, adversarial training and representation learning, and takes a step towards unsupervised conditional GANs. In particular, our approach exploits the structure of a latent space (learned by the representation learning) and employs it to condition the generative model. In this way, we break the traditional dependency between condition and label, substituting the latter by unsupervised features coming from the latent space. Finally, we show that this new technique is able to produce samples on demand keeping the quality of its supervised counterpart.

Download


Paper Citation


in Harvard Style

Durall R., Ho K., Pfreundt F. and Keuper J. (2021). Latent Space Conditioning on Generative Adversarial Networks.In Proceedings of the 16th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 4: VISAPP, ISBN 978-989-758-488-6, pages 24-34. DOI: 10.5220/0010178800240034


in Bibtex Style

@conference{visapp21,
author={Ricard Durall and Kalun Ho and Franz-Josef Pfreundt and Janis Keuper},
title={Latent Space Conditioning on Generative Adversarial Networks},
booktitle={Proceedings of the 16th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 4: VISAPP,},
year={2021},
pages={24-34},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0010178800240034},
isbn={978-989-758-488-6},
}


in EndNote Style

TY - CONF

JO - Proceedings of the 16th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 4: VISAPP,
TI - Latent Space Conditioning on Generative Adversarial Networks
SN - 978-989-758-488-6
AU - Durall R.
AU - Ho K.
AU - Pfreundt F.
AU - Keuper J.
PY - 2021
SP - 24
EP - 34
DO - 10.5220/0010178800240034