Realtime Multi-Person 2D Pose Estimation using Part Affinity Fields ∗

1. 只找一个limb,即只考虑a single pair of parts j1 and j2, 此时对应到：带权重二分图的匹配问题，权重就是积分出来的E，目标就是找到累积权重最大的匹配
2. When it comes to finding the full body pose of multiple
   people, determining Z is a K-dimensional matching problem.
   This problem is NP Hard [32] and many relaxations
   exist. In this work, we add two relaxations to the optimization,
   specialized to our domain. First, we choose a minimal
   number of edges to obtain a spanning tree skeleton of human
   pose rather than using the complete graph, as shown in
   Fig. 6c. Second, we further decompose the matching problem
   into a set of bipartite matching subproblems and determine
   the matching in adjacent tree nodes independently,
   as shown in Fig. 6d.
3. Our optimization scheme over
   the tree structure is orders of magnitude faster than the optimization
   over the fully connected graph 效率提高了几个数量级
4. PCKh: 各个part落在groundtruth附近，若在head size之内，认为检测出了。PCK的阈值是max(height,width) of body 的0.1 or 0.2倍。同时还可以分析PCKh-0.5等等精度的准确率
5. mAP
6. OKS: object keypoint similarity: It is calculated from scale of the person and the
   distance between predicted points and GT points.
7. It is noteworthy
   that our method has lower accuracy than the top-down
   methods on people of smaller scales (APM). The reason is
   that our method has to deal with a much larger scale range
   spanned by all people in the image in one shot. In contrast,
   top-down methods can rescale the patch of each detected
   area to a larger size and thus suffer less degradation
   at smaller scales.
8. If we use the GT bounding box and a single
   person CPM [31], we can achieve a upper-bound for
   the top-down approach using CPM, which is 62.7% AP.
   If we use the state-of-the-art object detector, Single Shot
   MultiBox Detector (SSD)[16], the performance drops 10%.
   This comparison indicates the performance of top-down approaches
   rely heavily on the person detector就是说自顶向下的方法每次都处理一个人，需要有个bounding box把人框出来(1. person detection 2. CPM或其他单人姿态检测算法)，然后对框出来的单个人图像进行尺度调整，到一个合适的图像大小，再进行处理。当bounding box精度不够时，自顶向下的方法误差会很高。
9.