EUVIMEDEUVIMED entdecken
Crossref

Evolutionary HyperNet-InfoGAN Baselines for Controllable Synthetic Image Generation

Mark Nuppnau, Khalid Kattan, R. G. Reynolds

Proceedings of the AAAI Symposium Series · 18.05.2026 · 8 (1) · 723-727

DOI: 10.1609/aaaiss.v8i1.42612 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Proceedings of the AAAI Symposium Series, online erschienen am 18.05.2026. Verfasst von Mark Nuppnau und 2 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Synthetic medical data generation requires human-auditable control. It is important that latent factors are able to separate data-specific structures so clinicians can inspect, trust, and responsibly reuse synthetic samples. Building on prior evolutionary generator work, an Evolutionary HyperNet-InfoGAN is developed in which Policy Gradients with Parameter-Based Exploration (PGPE) optimizes a compact HyperNetwork that emits the weights of a larger image generator, while a discriminator/Qnetwork is trained with backpropagation. This process creates a non-stationary adversarial setting in which evolutionary search must continually adapt, but in a far smaller and more structured parameter space than direct evolution of the full generator. This model is used to evaluate a shared baseline on the MNIST dataset (28x28 grayscale, 10 classes) and then on a grayscale-converted BloodMNIST dataset (28x28, 8 classes) with a shared-z protocol in which groups of samples share continuous noise while discrete codes vary. This setup enables the direct measurement of feature-space code separation (rsense) and within-code variation (rintra). When compared to a previous PGPE baseline that searched directly over a generator with more than 500k parameters, the HyperNetwork formulation yielded cleaner code allocation, substantially more stable late-stage training, and the ability to resolve early duplicate modes to distinct digits for better mode coverage. When applied to MNIST, the baseline approached full digit coverage with fewer duplicate-code failures than prior direct search. For BloodMNIST, the preliminary baseline results show partial disentanglement and class-consistent morphology, suggesting that HyperNetworks provide a practical intermediate step toward auditable, controllable synthetic medical image generation. These baselines motivate the next stage of this work that will employ a Cultural Algorithms (CA) as a meta-level controller in order to support adaptive fitness weighting and secondary search guidance within the PGPE framework.

Zitieren

APA

Nuppnau, M., Kattan, K., & Reynolds, R. G. (2026). Evolutionary HyperNet-InfoGAN Baselines for Controllable Synthetic Image Generation. Proceedings of the AAAI Symposium Series, 8(1), 723-727. https://doi.org/10.1609/aaaiss.v8i1.42612

Vancouver

Nuppnau M, Kattan K, Reynolds RG. Evolutionary HyperNet-InfoGAN Baselines for Controllable Synthetic Image Generation. Proceedings of the AAAI Symposium Series. 2026;8(1):723-727. doi:10.1609/aaaiss.v8i1.42612

BibTeX
@article{Nuppnau2026,
  author = {Nuppnau, Mark and Kattan, Khalid and Reynolds, R. G.},
  title = {Evolutionary HyperNet-InfoGAN Baselines for Controllable Synthetic Image Generation},
  journal = {Proceedings of the AAAI Symposium Series},
  year = {2026},
  volume = {8},
  number = {1},
  pages = {723--727},
  doi = {10.1609/aaaiss.v8i1.42612}
}
Kostenlos starten →