Research

Paper

TESTING March 18, 2026

DSS-GAN: Directional State Space GAN with Mamba backbone for Class-Conditional Image Synthesis

Authors

Aleksander Ogonowski, Konrad Klimaszewski, Przemysław Rokita

Abstract

We present DSS-GAN, the first generative adversarial network to employ Mamba as a hierarchical generator backbone for noise-to-image synthesis. The central contribution is Directional Latent Routing (DLR), a novel conditioning mechanism that decomposes the latent vector into direction-specific subvectors, each jointly projected with a class embedding to produce a feature-wise affine modulation of the corresponding Mamba scan. Unlike conventional class conditioning that injects a global signal, DLR couples class identity and latent structure along distinct spatial axes of the feature map, applied consistently across all generative scales. DSS-GAN achieves improved FID, KID, and precision-recall scores compared to StyleGAN2-ADA across multiple tested datasets. Analysis of the latent space reveals that directional subvectors exhibit measurable specialization: perturbations along individual components produce structured, direction-correlated changes in the synthesized image.

Metadata

arXiv ID: 2603.17637
Provider: ARXIV
Primary Category: cs.LG
Published: 2026-03-18
Fetched: 2026-03-19 06:01

Related papers

Raw Data (Debug)
{
  "raw_xml": "<entry>\n    <id>http://arxiv.org/abs/2603.17637v1</id>\n    <title>DSS-GAN: Directional State Space GAN with Mamba backbone for Class-Conditional Image Synthesis</title>\n    <updated>2026-03-18T11:58:47Z</updated>\n    <link href='https://arxiv.org/abs/2603.17637v1' rel='alternate' type='text/html'/>\n    <link href='https://arxiv.org/pdf/2603.17637v1' rel='related' title='pdf' type='application/pdf'/>\n    <summary>We present DSS-GAN, the first generative adversarial network to employ Mamba as a hierarchical generator backbone for noise-to-image synthesis. The central contribution is Directional Latent Routing (DLR), a novel conditioning mechanism that decomposes the latent vector into direction-specific subvectors, each jointly projected with a class embedding to produce a feature-wise affine modulation of the corresponding Mamba scan. Unlike conventional class conditioning that injects a global signal, DLR couples class identity and latent structure along distinct spatial axes of the feature map, applied consistently across all generative scales. DSS-GAN achieves improved FID, KID, and precision-recall scores compared to StyleGAN2-ADA across multiple tested datasets. Analysis of the latent space reveals that directional subvectors exhibit measurable specialization: perturbations along individual components produce structured, direction-correlated changes in the synthesized image.</summary>\n    <category scheme='http://arxiv.org/schemas/atom' term='cs.LG'/>\n    <category scheme='http://arxiv.org/schemas/atom' term='cs.CV'/>\n    <published>2026-03-18T11:58:47Z</published>\n    <arxiv:primary_category term='cs.LG'/>\n    <author>\n      <name>Aleksander Ogonowski</name>\n    </author>\n    <author>\n      <name>Konrad Klimaszewski</name>\n    </author>\n    <author>\n      <name>Przemysław Rokita</name>\n    </author>\n  </entry>"
}