论文标题
重新访问密集对象检测的AP损失：自适应排名对选择
Revisiting AP Loss for Dense Object Detection: Adaptive Ranking Pair Selection
论文作者
论文摘要
平均精度（AP）损失最近在密集的对象检测任务上显示出有希望的性能。但是，对AP损失如何从成对排名的角度影响检测器的深刻了解尚未开发出来。在这项工作中，我们重新审视了平均精度（AP）损失，并揭示了至关重要的元素是选择正面和负样本之间的排名对。基于这一观察结果，我们提出了两种策略，以改善AP损失。第一个是一种新型的自适应成对误差（APE）损失，该损失重点是正面和负样本中的排名对。此外，我们通过使用聚类算法利用归一化排名得分和本地化得分来选择更准确的排名对。在MSCOCO数据集上进行的实验支持我们的分析，并证明了我们提出的方法的优越性与当前分类和排名损失相比。该代码可从https://github.com/xudangliatiger/ape-loss获得。
Average precision (AP) loss has recently shown promising performance on the dense object detection task. However,a deep understanding of how AP loss affects the detector from a pairwise ranking perspective has not yet been developed.In this work, we revisit the average precision (AP)loss and reveal that the crucial element is that of selecting the ranking pairs between positive and negative samples.Based on this observation, we propose two strategies to improve the AP loss. The first of these is a novel Adaptive Pairwise Error (APE) loss that focusing on ranking pairs in both positive and negative samples. Moreover,we select more accurate ranking pairs by exploiting the normalized ranking scores and localization scores with a clustering algorithm. Experiments conducted on the MSCOCO dataset support our analysis and demonstrate the superiority of our proposed method compared with current classification and ranking loss. The code is available at https://github.com/Xudangliatiger/APE-Loss.
