针对细长结构分割的几何失配问题，提出频率感知的定向序列化方法。
Bridging the Geometry Mismatch: Frequency-Aware Anisotropic Serialization for Thin-Structure SSMs
- 通过频域解耦分离拓扑与高频方向特征，修正下采样导致的空间错位。
- 采用频率对齐扫描策略，在4个基准上达到91.3% mIoU和97.1% clDice。
- 适合处理裂纹、血管等细长结构分割，兼具高精度与实时性（80 FPS）
细长线状结构的分割具有拓扑敏感特性，局部微小误差即可能导致长程连接断裂。尽管近期状态空间模型（SSMs）具备高效长程建模能力，但其各向同性序列化（如栅格扫描）与各向异性目标存在几何失配，导致状态传播偏离结构轨迹而非沿其延伸。为此，本文提出FGOS-Net框架，基于频率-几何解耦思想，将特征分解为稳定拓扑载体与方向性高频分量，利用后者显式校正下采样引起的空间错位。在此校准拓扑基础上，引入频率对齐扫描，使序列化过程成为受几何条件约束的决策行为，从而保持方向一致的轨迹追踪。结合主动探测策略，选择性注入高频细节并抑制纹理歧义，FGOS-Net在四个挑战性基准上持续优于强基线。特别地，在DeepCrack数据集上达到91.3% mIoU与97.1% clDice，且仅需7.87 GFLOPs，推理速度达80 FPS。
原文摘要 · Abstract (English)
The segmentation of thin linear structures is inherently topology allowbreak-critical, where minor local errors can sever long-range connectivity. While recent State-Space Models (SSMs) offer efficient long-range modeling, their isotropic serialization (e.g., raster scanning) creates a geometry mismatch for anisotropic targets, causing state propagation across rather than along the structure trajectories. To address this, we propose FGOS-Net, a framework based on frequency allowbreak-geometric disentanglement. We first decompose features into a stable topology carrier and directional high-frequency bands, leveraging the latter to explicitly correct spatial misalignments induced by downsampling. Building on this calibrated topology, we introduce frequency-aligned scanning that elevates serialization to a geometry-conditioned decision, preserving direction-consistent traces. Coupled with an active probing strategy to selectively inject high-frequency details and suppress texture ambiguity, FGOS-Net consistently outperforms strong baselines across four challenging benchmarks. Notably, it achieves 91.3% mIoU and 97.1% clDice on DeepCrack while running at 80 FPS with only 7.87 GFLOPs.
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。