Probabilistic two-stage detection
Xingyi Zhou Vladlen Koltun Philipp Krähenbühl
Abstract
We develop a probabilistic interpretation of two-stage object detection. We show that this probabilistic interpretation motivates a number of common empirical training practices. It also suggests changes to two-stage detection pipelines. Specifically, the first stage should infer proper object-vs-background likelihoods, which should then inform the overall score of the detector. A standard region proposal network (RPN) cannot infer this likelihood sufficiently well, but many one-stage detectors can. We show how to build a probabilistic two-stage detector from any state-of-the-art one-stage detector. The resulting detectors are faster and more accurate than both their one- and two-stage precursors. Our detector achieves 56.4 mAP on COCO test-dev with single-scale testing, outperforming all published results. Using a lightweight backbone, our detector achieves 49.2 mAP on COCO at 33 fps on a Titan Xp, outperforming the popular YOLOv4 model.
中文速览
目标检测领域长期存在一个尴尬：两阶段检测器（先用区域提议网络RPN筛候选框、再分类）的两个阶段是各自独立训练的，缺乏统一的概率解释，导致最终得分只来自第二阶段，忽视了第一阶段的信息。这篇论文从概率论出发，把两阶段检测重新推导为一个联合概率模型——第一阶段估计"这里有没有物体"的可能性，第二阶段估计"是什么类别"的条件概率，两者相乘才是最终得分；同时推导出一个可高效优化的下界目标函数，让两个阶段协同训练。关键发现是：传统RPN为了追求高召回率而标注过于宽松，根本无法给出准确的"有物体"概率，但现成的单阶段检测器（如CenterNet）可以胜任这个角色——于是作者直接把单阶段检测器嵌入为第一阶段，再接一个级联分类头作为第二阶段，构建出"概率两阶段检测器"CenterNet2。实验结果表明，这一框架在COCO数据集上以单尺度测试达到56.4 mAP，超越所有已发表方法，轻量版本在Titan Xp上以33帧每秒跑出49.2 mAP，显著优于YOLOv4，兼顾了精度与速度，也为如何在理论层面统一两阶段检测提供了清晰的框架。
原文 arXiv:2103.07461;中英对照 + 大白话阅读 https://aha.fim.ai/paper/2103.07461v1