SetPlanner：面向冻结SAM的轻量级即插即用点集规划器
SetPlanner: A Lightweight Plug-in Point-Set Planner for Frozen SAM
AI总结：
针对冻结SAM的自动提示难题，提出1.52M参数的即插即用点集规划器SetPlanner，通过置换感知条件流规划点集，在三个内窥镜数据集上全面超越LoRA适配系统，Dice达0.934。
AI中文摘要：
Segment Anything Models提供了可复用的先验，但它们需要用户提示，无法支持全自动的器械分割。对于细长、关节式、反光且部分遮挡的工具，自动提示较为困难，因为多种配置可能都是有效的。我们将自动提示问题形式化为轻量级点集规划，并在冻结路径下隔离点源。为此，我们提出了SetPlanner，一个参数量为1.52M的即插即用点集规划器，用于冻结的SAM。该插件保留了SAM的点提示接口，并支持在不同骨干网络间复用。SetPlanner通过置换感知的条件流，从几何感知目标中规划完整的无序K点集。SAM解码出八个候选结果；它们的共识读出产生一个无需真实标签的预测。在三个内窥镜数据集上，SetPlanner在全部六条迁移路径上均优于基于LoRA适配的系统。在我们的冻结路径协议下，SetPlanner在Kvasir-Instrument上达到0.934的Dice系数，并恢复了44.4点定位差距的96%，而候选分歧在AUROC为0.969的情况下对低Dice案例进行排序。
英文摘要：
Segment Anything Models provide reusable priors, yet they require user prompts and cannot support fully automatic instrument segmentation. Automatic prompting is difficult for thin, articulated, reflective, and partly occluded tools, where several configurations can be valid. We formulate automatic prompting as lightweight point-set planning and isolate the point source under a frozen pathway. To this end, we present SetPlanner, a 1.52M-parameter plug-in point-set planner for frozen SAM. The plug-in preserves SAM's point-prompt interface and enables reuse across backbones. SetPlanner plans complete unordered K-point sets from geometry-aware targets with a permutation-aware conditional flow. SAM decodes eight candidates; their consensus readout yields a ground-truth-free prediction. Across three endoscopic datasets, SetPlanner wins all six transfer routes over a LoRA-adapted system. Under our frozen-pathway protocol, SetPlanner reaches 0.934 Dice on Kvasir-Instrument and recovers 96% of a 44.4-point localization gap, while candidate disagreement ranks low-Dice cases at AUROC 0.969.
