毕设终于告一段落，传统方法的视觉做得我整个人都很奔溃，终于结束，可以看些搁置很久的一些论文了，嘤嘤嘤
Densely Connected Convolutional Networks 其实很早就出来了，cvpr 2017 best paper
觉得读论文前，还是把dense net的整个网络结构放到http://ethereon.github.io/netscope/#/editor 上面可视化看一下，会更加容易理解,总体这篇论文很好理解
上图是一个5层的dense block,每个dense block的growth rate k=4
论文开头给出了densnet的几个优点：
1、Our proposed DenseNet architecture explicitly differentiates between information that is added to the network and information that is preserved.DenseNet layers are very narrow (e.g., 12 filters per layer),adding only a small set of feature-maps to the “collective knowledge” of the network and keep the remaining featuremaps
unchanged—and the final classifier makes a decision based on all feature-maps in the networkdensnet 网络结构参数少，每个block里面的filter也比较少，而我们在使用alexnet，通常filter都是上百的，而这里的filter 12、24、16 等，所以非常narrow
2、one big advantage of DenseNets is their improved flow of information and gradients throughout the network, which makes them easy to train.
densenet网络从上图中可以看出，每层都和后面的层相连，（第一幅图没有画出来每个block中的层之间的连接，觉得应该结合第一个图和第二个图，才算完整，因为第二个图每个block后面的输入是前面所有层concat一起的结果，相当于图一显示的那样。。。。在可视化工具里面看，最明显了，而且还能看到每一层的实际大小）有利于信息和梯度在整个网络中的传递。
3、we also observe that dense connections have a regularizing effect, which reduces overfitting on tasks with smaller training set sizes.
同时densenet网络也有正则化的作用，在小数据集上训练也能减少过拟合的风险
densenet中是将前面几层concat一起，而resnet是求和，论文中提到这种求和会影响信息在网络中的传递
transition layers:
这层就是连接两个block之间的层，由BN层，1x1 卷积层和2x2的avg pooling层构成，如下图所示
Growth rate:
也就是每个block里面的层数，如图一中，每个block里面有4层，所以growth rate=4
Botttleneck layers:
It has been noted in [36, 11] that a 1X1 convolution can be introduced as bottleneck layer before each 3X3 convolution to reduce the number of input feature-maps, and thus to improve computational efficiency.
是指每一层中在3X3卷积核前面有个1X1的卷积核，作用是减少输入feature-map的数量，如下图所示，512的数量变成了128个
Compression:
If a dense block contains m feature-maps, we let the following transition layer generate [θm] output featuremaps,where 0<θ <=1referred to as the compression factor.
让transition layer压缩block输出的feature map数量。
