用Mamba做生成器，通过方向性潜变量提升图像合成质量
DSS-GAN: Directional State Space GAN with Mamba backbone for Class-Conditional Image Synthesis
- 设计方向潜变量路由机制，将潜在向量按方向拆分并逐层调制
- 在多个数据集上超越StyleGAN2-ADA，FID、KID和精确率-召回率均更好
- 潜空间分析显示各方向分量具可解释性，扰动产生结构化图像变化
我们提出DSS-GAN，首个采用Mamba作为层级生成器骨干的生成对抗网络，用于噪声到图像的合成。核心贡献是方向潜变量路由（DLR），一种新型条件机制：将潜在向量分解为方向相关的子向量，每个子向量与类别嵌入联合投影，生成特征级仿射调制，作用于对应的Mamba扫描。与传统全局信号注入不同，DLR将类别身份与潜在结构沿特征图的不同空间轴耦合，并在所有生成尺度上保持一致。DSS-GAN在多个测试数据集上优于StyleGAN2-ADA，FID、KID及精确率-召回率均更优。潜空间分析表明，方向子向量具有显著特化性：沿单个分量扰动会引发图像中方向相关且结构化的改变。
原文摘要 · Abstract (English)
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.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。