文章目录
1、论文总述
这篇论文有点结合centerNet、ROIalign、DCN的意思，它已经不是像RPN、SSD那样提前预定义好许多的anchor，而是根据目标的一些语义信息先将目标的中心点学出来，然后根据中心点附近的特征回归出ROI的w、h，这样就能在feature map的每个点得到一个对应的anchor，然后利用feature adaption module（DCN操作）来将学出来的proposals与对应的feature对齐，然后进行分类回归，所以它更像是升级版一个成熟的RPN，故又叫GA-RPN，可以将其与onestage与twostage检测器相结合。
In this work, we present a more effective method to prepare anchors, with the aim to mitigate the issues of handpicked priors. Our method is motivated by the observation
that objects are not distributed evenly over the image. The
scale of an object is also closely related to the imagery content, its location and geometry of the scene. Following this
intuition, our method generates sparse anchors in two steps:
first identifying sub-regions that may contain objects and
then determining the shapes at different locations.
Learnable anchor shapes are promising, but it breaks the
aforementioned rule of consistency, thus presents a new
challenge for learning anchor representations for accurate
classification and regression. Scales and aspect ratios of anchors are now variable instead of fixed, so different feature
map pixels have to learn adaptive representations that fit the
corresponding anchors. To solve this problem, we introduce
an effective module to adapt the features based on anchor
geometry.
解读这篇论文时，在知乎上看到原作者陈凯的解读，Guided Anchoring: 物体检测器也能自己学 Anchor，感觉这篇文章解读的已经很详细，所以论文具体内容就不自己写了，请参考它的。
【注】：该算法中心点固定，回归时只回归shape。
2、anchor设计的两条准则
There are two general rules for a reasonable anchor de-
sign: alignment and consistency.
Firstly, to use convolutional features as anchor representations, anchor centers need to be well aligned with feature map pixels.
Secondly, the receptive field and semantic scope should be consistent
with the scale and shape of anchors on different locations of
a feature map。
参考文献
