用轻量拓扑感知适配器，仅调5.2%参数就让SAM精准分割细结构
TopoLoRA-SAM: Topology-Aware Parameter-Efficient Adaptation of Foundation Segmenters for Thin-Structure and Cross-Domain Binary Semantic Segmentation
- 在冻结ViT编码器中注入低秩适配模块，加轻量卷积适配器
- 仅需490万参数（5.2%）训练，平均Dice达最优，挑战数据上提升显著
- 适合医疗/遥感等细结构与噪声模态场景，无需全模型微调
基础分割模型如段一切模型（SAM）通过大规模预训练展现出强大的零样本泛化能力，但将其适配到特定领域语义分割仍具挑战性，尤其在细结构（如视网膜血管）和噪声模态（如合成孔径雷达影像）下。全量微调计算开销大且易导致灾难性遗忘。我们提出TopoLoRA-SAM，一种面向二值语义分割的拓扑感知、参数高效的适配框架。该方法将低秩适配（LoRA）注入冻结的ViT编码器，并引入轻量级空间卷积适配器，可选地通过可微分clDice实现拓扑感知监督。我们在五个基准上评估：视网膜血管分割（DRIVE、STARE、CHASE_DB1）、息肉分割（Kvasir-SEG）及SAR海陆分割（SL-SSDD），对比U-Net、DeepLabV3+、SegFormer和Mask2Former。TopoLoRA-SAM在视网膜数据集平均Dice和跨数据集整体平均Dice均表现最佳，训练参数仅为模型总量的5.2%（约490万）。在具有挑战性的CHASE_DB1数据集上，本方法显著提升分割精度与鲁棒性，表明拓扑感知的参数高效适配可媲美甚至超越全微调的专业模型。代码已开源：https://github.com/salimkhazem/Seglab.git
原文摘要 · Abstract (English)
Foundation segmentation models such as the Segment Anything Model (SAM) exhibit strong zero-shot generalization through large-scale pretraining, but adapting them to domain-specific semantic segmentation remains challenging, particularly for thin structures (e.g., retinal vessels) and noisy modalities (e.g., SAR imagery). Full fine-tuning is computationally expensive and risks catastrophic forgetting. We propose \textbf{TopoLoRA-SAM}, a topology-aware and parameter-efficient adaptation framework for binary semantic segmentation. TopoLoRA-SAM injects Low-Rank Adaptation (LoRA) into the frozen ViT encoder, augmented with a lightweight spatial convolutional adapter and optional topology-aware supervision via differentiable clDice. We evaluate our approach on five benchmarks spanning retinal vessel segmentation (DRIVE, STARE, CHASE\_DB1), polyp segmentation (Kvasir-SEG), and SAR sea/land segmentation (SL-SSDD), comparing against U-Net, DeepLabV3+, SegFormer, and Mask2Former. TopoLoRA-SAM achieves the best retina-average Dice and the best overall average Dice across datasets, while training only \textbf{5.2\%} of model parameters ($\sim$4.9M). On the challenging CHASE\_DB1 dataset, our method substantially improves segmentation accuracy and robustness, demonstrating that topology-aware parameter-efficient adaptation can match or exceed fully fine-tuned specialist models. Code is available at : https://github.com/salimkhazem/Seglab.git
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。