Abstract
Conditional Generative Adversarial Networks (CGANs) are diversely utilized for data synthesis in applied sciences and natural image tasks. Conditional generative models extend upon data generation to account for labeled data by estimating joint distributions of samples and labels. We present a family of modified CGANs which demonstrate the inclusion of reconstructive cycles between prior and data spaces inspired by BiGAN and CycleGAN improves upon baselines for natural image synthesis with three primary contributions. The first is a study proposing three incremental architectures for conditional data generation which demonstrate improvement on baseline generation quality for a natural image data set across multiple generative metrics. The second is a novel approach to structure latent representations by learning a paired structured condition space and weakly structured variation space with desirable sampling and supervised learning properties. The third is a proposed utilization of conditional image synthesis for supervised learner data set augmentation as an alternative generation metric. Additional experiments demonstrate the successes of inducing cycles in conditional GANs for both image synthesis and image classification over comparable models with no additional tweaks or modifications. We release our source code, models, and experiments here: https://github.com/alexander-moore/Cycles-Improve-Conditional-Generators.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Brock, A., Donahue, J., Simonyan, K.: Large scale gan training for high fidelity natural image synthesis (2018). https://doi.org/10.48550/ARXIV.1809.11096, https://arxiv.org/abs/1809.11096
Child, R.: Very deep vaes generalize autoregressive models and can outperform them on images (2021)
Dai, B., Lin, D.: Contrastive learning for image captioning. arXiv preprint arXiv:1710.02534 (2017)
DeVries, T., Romero, A., Pineda, L., Taylor, G.W., Drozdzal, M.: On the evaluation of conditional gans. arXiv preprint arXiv:1907.08175 (2019)
Donahue, J., Krähenbühl, P., Darrell, T.: Adversarial feature learning. arXiv preprint arXiv:1605.09782 (2016)
Goodfellow, I.J., et al.: Generative adversarial networks. arXiv preprint arXiv:1406.2661 (2014)
Heusel, M., Ramsauer, H., Unterthiner, T., Nessler, B., Hochreiter, S.: Gans trained by a two time-scale update rule converge to a local nash equilibrium. arXiv preprint arXiv:1706.08500 (2017)
Karras, T., Aittala, M., Hellsten, J., Laine, S., Lehtinen, J., Aila, T.: Training generative adversarial networks with limited data. arXiv preprint arXiv:2006.06676 (2020)
Kingma, D.P., Welling, M.: Auto-encoding variational bayes (2014)
Krizhevsky, A., Hinton, G., et al.: Learning multiple layers of features from tiny images (2009)
Mirza, M., Osindero, S.: Conditional generative adversarial nets (2014)
Odena, A., Olah, C., Shlens, J.: Conditional image synthesis with auxiliary classifier gans (2017)
Radford, A., Metz, L., Chintala, S.: Unsupervised representation learning with deep convolutional generative adversarial networks. arXiv preprint arXiv:1511.06434 (2015)
Sauer, A., Schwarz, K., Geiger, A.: Stylegan-xl: scaling stylegan to large diverse datasets (2022). https://doi.org/10.48550/ARXIV.2202.00273, https://arxiv.org/abs/2202.00273
Suman Ravuri, O.V.: Seeing is not necessarily believing: limitations of biggans for data augmentation. In: International Conference on Learning Representations Workshop 2019 (2019)
Szegedy, C., Vanhoucke, V., Ioffe, S., Shlens, J., Wojna, Z.: Rethinking the inception architecture for computer vision (2015)
Zhang, H., Zhang, Z., Odena, A., Lee, H.: Consistency regularization for generative adversarial networks. arXiv preprint arXiv:1910.12027 (2019)
Zhu, J.Y., Park, T., Isola, P., Efros, A.A.: Unpaired image-to-image translation using cycle-consistent adversarial networks. In: Proceedings of the IEEE International Conference on Computer Vision, pp. 2223–2232 (2017)
Acknowledgements
This manuscript has been authored with funding provided by the Defense Threat Reduction Agency (DTRA). The publisher acknowledges that the US Government retains a nonexclusive, paid-up, irrevocable, worldwide license to publish or reproduce the published form of this manuscript, or allow others to do so, for US government purposes. Approved for public release. Distribution unlimited.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2022 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this paper
Cite this paper
Moore, A.M., Paffenroth, R.C., Ngo, K.T., Uzarski, J.R. (2022). Cycles Improve Conditional Generators: Synthesis and Augmentation for Data Mining. In: Chen, W., Yao, L., Cai, T., Pan, S., Shen, T., Li, X. (eds) Advanced Data Mining and Applications. ADMA 2022. Lecture Notes in Computer Science(), vol 13726. Springer, Cham. https://doi.org/10.1007/978-3-031-22137-8_26
Download citation
DOI: https://doi.org/10.1007/978-3-031-22137-8_26
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-031-22136-1
Online ISBN: 978-3-031-22137-8
eBook Packages: Computer ScienceComputer Science (R0)