问题：Deep convolutional neural networks have greatly improved object recognition accuracy,but remain reliant on large quantities of labeled training data.（深度卷积神经网络依赖大量标注的训练集）
目的：reduce annotation costs associated with detection.（减少目标检测的标注成本）
方法：unsupervised adaptation method for object detection that combines weak global alignment with strong local alignment.
（结合弱全局对准和强局部对准的非监督自适应目标检测模型）
具体做法：We extract global features just before the RPN and local features from lower layers, and perform weak global alignment in the high-level feature space and strong local alignment in the low-level feature space. （在RPN之前提取全局特征，在较低层提取局部特征，在高层特征空间进行弱全局对准操作，在低层特征空间进行强局部对准操作）
Weak Global Feature Alignment：
We propose to train a domain classifier to ignore easy-to-classify examples while focusing on hard-to-classify examples with respect to the classification of the domain.（我们提出训练一个忽略容易分类的样本，而专注于相对于域的分类难分类的样本的域分类器）
Strong Local Feature Alignment：
The architecture of the local domain classifier, Dl, is designed to focus on the local features rather than global features. Dl is a fully-convolutional network with kernel-size equal to one. （局部域分类器专注于局部特征而不是全局特征，局部域分类器是尺寸为1的全卷积网络。）
理解：用一个分类器区分背景和目标。。。