发表机构
School of Mechatronics Engineering, Harbin Institute of Technology（哈尔滨工业大学机电工程学院）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本文提出GeoDistill-Refine两阶段框架，将SAM 3伪掩码迁移至紧凑分割网络，通过多提示投票、样本级门控及多几何目标优化，提升航天器图像分割精度，模型高效且适配多域评估。
AI 中文摘要
基础分割模型可为航天器图像提供无需人工训练掩码的监督，但其预测会随文本提示变化，且可能存在几何误差，这类误差在蒸馏过程中会被放大。本文提出GeoDistill-Refine，这是一个两阶段框架，用于将离线SAM 3伪掩码迁移至紧凑的分割网络。6个固定提示通过无加权50%投票融合，以稳定教师模型输出。学生模型首先学习前景轮廓，随后通过从伪掩码衍生的有符号距离场、骨架和面积目标进行优化。由提示一致性、有效提示比例和伪掩码面积合理性计算得到的样本级门控，可降低不可靠伪几何的影响。在SpaceSense-Bench HJM锁箱集上，GeoDistill-Refine相比普通伪标签学生模型，将图像交并比（Image IoU）和边界F1值分别提升0.0456和0.1380。在SPEED+ Lightbox、Sunlamp域及TANGO上的外部评估显示，该方法在实现具有竞争力的区域重叠的同时，还提升了边界质量或前景精度。部署的TinyUNet包含0.263M参数，在RTX 4090上每张图像推理耗时约1.1ms；SAM 3伪掩码构建及辅助几何分支仅在训练阶段使用。
英文摘要
Foundation segmentation models can provide supervision for spacecraft imagery without manual training masks, but their predictions vary with textual prompts and may contain geometric errors that are amplified during distillation. This paper presents GeoDistill-Refine, a two-stage framework that transfers offline SAM 3 pseudo-masks to a compact segmentation network. Six fixed prompts are fused by an unweighted 50% vote to stabilize the teacher output. The student first learns the foreground silhouette and is then refined with signed-distance-field, skeleton, and area objectives derived from the pseudo-mask. A sample-level gate, computed from prompt agreement, the valid-prompt ratio, and pseudo-mask area plausibility, reduces the influence of unreliable pseudo-geometry. On the SpaceSense-Bench HJM lockbox set, GeoDistill-Refine improves Image IoU and Boundary F1 by 0.0456 and 0.1380, respectively, over a plain pseudo-label student. External evaluations on the SPEED+ Lightbox and Sunlamp domains and on TANGO show competitive regional overlap together with gains in boundary quality or foreground precision. The deployed TinyUNet contains 0.263 M parameters and requires approximately 1.1 ms per image on an RTX 4090; SAM 3 pseudo-mask construction and the auxiliary geometry branches are used only during training.
