发表机构
Carnegie Mellon University（卡内基梅隆大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
CoSWA-YOLOv12通过尺度自适应Wasserstein分配、小波细节残差和M2-NWD损失，在保持预训练权重可迁移的同时，显著提升疟原虫微小目标检测与分割的召回率和定位精度。
AI 中文摘要
自动显微镜检查可以扩大在资源匮乏环境中获得疟疾诊断的机会，但最致命的物种恶性疟原虫（P. falciparum）在其早期环状体阶段仅呈现为宽约几十像素的目标。这类微小目标被系统性地漏检：基于重叠的标签分配使它们缺乏正样本，而基于重叠的边界框回归在其尺度上产生弱梯度。归一化高斯Wasserstein距离（NWD）修复了这两种效应，但在整张玻片上均匀应用时，该玻片还包含比其大三到四倍的目标，这会放松对这些较大目标的监督并削弱其定位精度，因此即使微小类别有所改善，整体准确率也可能下降。我们提出了CoSWA-YOLOv12，一个紧凑的YOLOv12实例分割检测器，其核心的协同尺度自适应Wasserstein分配（Cooperative Scale-adaptive Wasserstein Assignment）将Wasserstein处理按与目标尺寸成反比的方式路由给目标，对于较大的物种逐渐回退到标准分配。另外两个组件为其提供支持：一个基于小波的细节残差，以及一个最小-最大高斯回归损失（M2-NWD）。所有三项新增功能均具有迁移安全性：每个都在初始化时精确复现标准预训练模型，因此公共预训练权重可在不损失任何精度的情况下加载。在一个五类别的卢旺达厚血膜数据集上，CoSWA-YOLOv12将恶性疟原虫的召回率从0.63提高到0.74，mAP@50从0.73提高到0.81（掩膜），将漏检的恶性疟原虫从38%降至15%，并在所有五个类别上改善了严格定位的mAP@50-95（检测和分割均如此），而2x2消融实验表明尺度门控和回归损失具有协同作用。
英文摘要
Automated microscopy could widen access to malaria diagnosis in low-resource settings, but the deadliest species, P. falciparum, presents in its early ring stage as an object only a few tens of pixels wide. Such tiny targets are systematically under-detected: overlap-based label assignment starves them of positive samples, and overlap-based box regression gives weak gradients at their scale. The Normalized Gaussian Wasserstein Distance (NWD) repairs both effects, but applied uniformly across a slide that also holds objects three to four times larger it loosens their supervision and erodes their localisation, so overall accuracy can fall even as the tiny class improves. We present CoSWA-YOLOv12, a compact YOLOv12 instance-segmentation detector whose core Cooperative Scale-adaptive Wasserstein Assignment routes the Wasserstein treatment to an object in inverse proportion to its size, tapering back to standard assignment for larger species. Two further components support it: a wavelet detail residual, and a min-max Gaussian regression loss (M2-NWD). All three additions are transfer-safe: each reproduces the standard pretrained model exactly at initialisation, so public pretrained weights load without any loss of accuracy. On a five-class Rwandan thick-smear dataset, CoSWA-YOLOv12 raises P. falciparum recall from 0.63 to 0.74 and mAP@50 from 0.73 to 0.81 (mask), cuts missed P. falciparum from 38% to 15%, and improves strict-localisation mAP@50-95 on all five classes for both detection and segmentation, while a 2x2 ablation shows the scale gate and the regression loss are synergistic.
CommentsAccepted at The 3rd MIRASOL workshop, a satellite event at MICCAI 2026. To appear in Springer Lecture Notes in Computer Science (LNCS)
