论文阅读：Augmentation for small object detection

1、论文总述

本篇论文主要针对COCO数据集上的小目标性能进行改进，相对来说，是比较实用的数据处理技巧，适用于所有的anchor-free和anchor-based的目标检测模型。
（1）对含有小目标的图像进行过采样，一般过采样率取3时较好，这样大目标受到的影响不会太大，因为总的来说，类别不平衡、不同尺度间的目标数量不平衡这类不平衡问题都是数据造成的，和模型结构关系不大，而目前的CNN的训练方法又都是SGD，根据Loss来进行优化，所以就会导致哪类数据多，哪类数据的性能就好，例如大目标多， 那么大目标的性能就好，小狗类别相对猫类别的数据多，那模型对狗的检测能力就要好于猫。所以对待这种不平衡问题，从数据下手一般要简单高效更加实用些。
（2）对含有小目标的图像，对多个小目标进行复制-粘贴策略增加小目标在图像中的出现频率，这样训练时，小目标的正样本也会增多，让模型更加侧重于小目标的学习。但是这个粘贴不能多次随意粘贴，要保证粘贴的时候，不会与已有的标注目标产生重叠，而且粘贴次数小于3次，太多小目标的话会影响大目标的性能，这样会得不偿失。粘贴时候是根据COCO训练集里自带的mask信息，作者实验证明，粘贴完之后不需要用高斯模糊对边界进行处理，用了之后反而有点下降。

具体特别细的讲解，我就不写了，极市平台有一篇讲的比较好的文章，看它就够了。
小目标检测：数据增强

2、举例说明小目标的重要性

Small object detection is crucial in many downstream tasks. Detecting small
or distant objects in the high-resolution scene photographs from the car is necessary to deploy self-driving cars safely. Many objects, such as traffic signs [11,34] or pedestrians [31], are often barely visible on the high-resolution images. In
medical imaging, early detection of masses and tumors is crucial for making an
accurate, early diagnosis, when such elements can easily be only a few pixels in
size [3,29]. Automatic industrial inspection can also benefit from small object
detection by the localization of small defects that can be visible on the material
surfaces [1,30]. Another application is satellite image analysis, where objects,
such as cars, ships, and houses, must be effectively annotated [28,21]. With an
average of 0.5-5m per pixel resolution, these objects are again just a few pixels in size.

3、COCO数据集中小目标的分布情况

Total Object Area表示标注像素面积所占比例

4、Augmentation与Oversampling结合时的3种策略

In the second set of experiments, we investigate the effects of using augmentation
on small object detection and segmentation. We copy and paste all small objects
in each image once. We also oversample images with small objects to study the
interaction between the oversampling and augmentation strategies.
We test three settings. In the first setting, we replace each image with small
objects by the one with copy-pasted small objects. In the second setting, we
duplicate these augmented images to mimic oversampling. In the final setup, we
keep both the original images and augmented images, which is equivalent to oversampling the images with small objects by the factor of two, while augmenting
the duplicated copies with more small objects.

实验表明，第三种（原图和aug后的图都保留） 效果比较好。
如果只保留增强的图，那么模型会重点学习复制的小目标，因为这些合成的小目标与背景亮度、分布啥的不一致，学起来较容易，那么测试时效果就不怎么提升，但是测试集里如果也加上合成数据，那么在这个合成测试集上的效果也会提升。

5、复制-粘贴时的3种策略

There are different ways to copy-pasting small objects. We consider three separate strategies. First, we pick one small object in an image and copy-paste it
multiple times in random locations. Second, we choose numerous small objects
and copy-paste each of these exactly once in an arbitrary position. Lastly, we
copy-paste all small objects in each image multiple times in random places. In
all the cases, we use the third setting of augmentation above; that is, we keep
both the original image and its augmented copy

实验表明，将一张图上的多个小目标复制一次效果比其他两种方式好，因为复制太多会影响模型的大目标性能。

参考文献

小目标检测：数据增强