论文阅读：Probabilistic Anchor Assignment with IoU Prediction for Object Detection（PAA）

1、论文总述

本篇论文主要是针对anchor的正负样本分配做改进，动机可以看下图：

相比于其他工作，作者不仅在anchor assignment阶段提出了自适应分配正负样本的做法，而且在optimization阶段和后处理阶段也加入了想对应的改进，可谓很全面了。

当然，最主要的工作还是利用高斯混合模型对anchor进行自适应的正负样本分配，抛弃了原先的基于IOU的分配方法，基于IOU的其实只看了anchor的定位信息，但IOU大的定位虽然好一点，但并不一定就适合用来预测该目标（注：有可能包含了许多背景信息），还得看它的分类得分，这就是论文的出发点。

关于这篇论文有一篇比较全面的解读，看这个就够了，我就不重复写了：

ECCV2020：probabilistic anchor assignment (PAA)

With this motivation, we propose a probabilistic anchor assignment (PAA)
strategy that adaptively separates a set of anchors into positive and negative
samples for a GT box according to the learning status of the model associated
with it. To do so we first define a score of a detected bounding box that reflects
both the classification and localization qualities. We then identify the connection between this score and the training objectives and represent the score as the
combination of two loss objectives. Based on this scoring scheme, we calculate
the scores of individual anchors that reflect how the model finds useful cues to
detect a target object in each anchor. With these anchor scores, we aim to find
a probability distribution of two modalities that best represents the scores as
positive or negative samples as in Figure 1. Under the found probability distribution, anchors with probabilities from the positive component are high are
selected as positive samples. This transforms the anchor assignment problem to a
maximum likelihood estimation for a probability distribution where the parameters of the distribution is determined by anchor scores. Based on the assumption
that anchor scores calculated by the model are samples drawn from a probability distribution, it is expected that the model can infer the sample separation
in a probabilistic way, leading to easier training of the model compared to other
non-probabilistic assignments. Moreover, since positive samples are adaptively
selected based on the anchor score distribution, it does not require a pre-defined
number of positive samples nor an IoU threshold.
On top of that, we identify that in most modern object detectors, there
is inconsistency between the testing scheme (selecting boxes according to the
classification score only during NMS) and the training scheme (minimizing both
classification and localization losses). Ideally, the quality of detected boxes should
be measured based not only on classification but also on localization. To improve
this incomplete scoring scheme and at the same time to reduce the discrepancy
of objectives between the training and testing procedures, we propose to predict
the IoU of a detected box as a localization quality, and multiply the classification score by the IoU score as a metric to rank detected boxes. This scoring is
intuitive, and allows the box scoring scheme in the testing procedure to share
the same ground not only with the objectives used during training, but also with
the proposed anchor assignment strategy that brings both classification and localization into account, as depicted in Figure 2. Combined with the proposed
PAA, this simple extension significantly contributes to detection performance.
We also compare the IoU prediction with the centerness prediction [33, 40] and
show the superiority of the proposed method.
With an additional improvement in post-processing named score voting, each
of our methods shows clear improvements as revealed in the ablation studies.

2、 Anchor Assignment的研究进展和优缺点

The task of selecting which anchors (or locations for anchor-free models) are to be
designated as positive or negative samples has recently been identified as a crucial
factor that greatly affects a model’s performance [37, 40, 42]. In this regard,
several methods have been proposed to overcome the limitation of the IoUbased hard anchor assignment.
MetaAnchor [37] predicts the parameters of the
anchor functions (the last convolutional layers of detection heads) dynamically
and takes anchor shapes as an argument, which provides the ability to change
anchors in training and testing. Rather than enumerating pre-defined anchors
across spatial locations, GuidedAnchoring [34] defines the locations of anchors
near the center of GTs as positives and predicts their shapes. FreeAnchor [42]
proposes a detection-customized likelihood that considers both the recall and
precision of samples into account and determines positive anchors based on the
estimated likelihood. ATSS [40] suggests an adaptive anchor assignment that
calculates the mean and standard deviation of IoU values from a set of close
anchors for each GT. It assigns anchors whose IoU values are higher than the
sum of the mean and the standard deviation as positives. Although these works
show some improvements, they either require additional layers and complicated
structures [34, 37], or force only one anchor to have a full classification score
which is not desirable in cases where multiple anchors are of high quality and
competitive [42], or rely on IoUs between pre-defined anchors and GTs and
consider neither the actual content of the intersecting regions nor the model’s
learning status [40].
Similar to our work, MultipleAnchorLearning (MAL) [16] and NoisyAnchor [20] define anchor score functions based on classification and localization
losses. However, they do not model the anchor selection procedure as a likelihood maximization for a probability distribution; rather, they choose a fixed
number of best scoring anchors. Such a mechanism prevents these models from
selecting a flexible number of positive samples according to the model’s learning status and input. MAL uses a linear scheduling that reduces the number
of positives as training proceeds and requires a heuristic feature perturbation
to mitigate it. NoisyAnchor fixes the number of positive samples throughout
training. Also, they either miss the relation between the anchor scoring scheme
and the box selection objective in NMS [16] or only indirectly relate them using
soft-labels [20].

参考文献

1、ECCV2020：probabilistic anchor assignment (PAA)

2、（三十七）论文阅读 | 目标检测之PAA