X-GAN: Generative Adversarial Networks Training Guided with Explainable Artificial Intelligence

Guilherme Rozendo, Guilherme Rozendo, Alessandra Lumini, Guilherme Roberto, Thaína Tosta, Marcelo Zanchetta do Nascimento, Leandro Neves

2024

Abstract

Generative Adversarial Networks (GANs) create artificial images through adversary training between a generator (G) and a discriminator (D) network. This training is based on game theory and aims to reach an equilibrium between the networks. However, this equilibrium is hardly achieved, and D tends to be more powerful. This problem occurs because G is trained based on only a single value representing D’s prediction, and only D has access to the image features. To address this issue, we introduce a new approach using Explainable Artificial Intelligence (XAI) methods to guide the G training. Our strategy identifies critical image features learned by D and transfers this knowledge to G. We have modified the loss function to propagate a matrix of XAI explanations instead of only a single error value. We show through quantitative analysis that our approach can enrich the training and promote improved quality and more variability in the artificial images. For instance, it was possible to obtain an increase of up to 37.8% in the quality of the artificial images from the MNIST dataset, with up to 4.94% more variability when compared to traditional methods.

Paper Citation

in Harvard Style

Rozendo G., Lumini A., Roberto G., Tosta T., Zanchetta do Nascimento M. and Neves L. (2024). X-GAN: Generative Adversarial Networks Training Guided with Explainable Artificial Intelligence. In Proceedings of the 26th International Conference on Enterprise Information Systems - Volume 1: ICEIS; ISBN 978-989-758-692-7, SciTePress, pages 674-681. DOI: 10.5220/0012618400003690

in Bibtex Style

@conference{iceis24,
author={Guilherme Rozendo and Alessandra Lumini and Guilherme Roberto and Thaína Tosta and Marcelo Zanchetta do Nascimento and Leandro Neves},
title={X-GAN: Generative Adversarial Networks Training Guided with Explainable Artificial Intelligence},
booktitle={Proceedings of the 26th International Conference on Enterprise Information Systems - Volume 1: ICEIS},
year={2024},
pages={674-681},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0012618400003690},
isbn={978-989-758-692-7},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 26th International Conference on Enterprise Information Systems - Volume 1: ICEIS
TI - X-GAN: Generative Adversarial Networks Training Guided with Explainable Artificial Intelligence
SN - 978-989-758-692-7
AU - Rozendo G.
AU - Lumini A.
AU - Roberto G.
AU - Tosta T.
AU - Zanchetta do Nascimento M.
AU - Neves L.
PY - 2024
SP - 674
EP - 681
DO - 10.5220/0012618400003690
PB - SciTePress