UNet++
2018 CVPR
周纵苇,发表论文时是一个在读二年级的博士,在亚利桑那州立大学念生物信息学。
UNet++: A Nested U-Net Architecture for Medical Image Segmentation
1、Introduce
UNet++与U-Net的区别:
- Re-designed skip pathways
- Deep supervision
2、Network
- (a)黑色表示原始的Unet,绿色和蓝色表示skip pathways 上密集的卷积块,红色表示deep supervision
- (b) UNet++的第一个skip pathway 的详细分析。
- (c) 如果经过deep supervision训练,UNet++可以在推理时进行修剪。
Re-designed skip pathways
在U-Net中,编码器的特征图在解码器中直接接收;然而,在UNet++中,它们经历了一个密集的卷积块,其卷积层数依赖于金字塔级。
:denote the output of node
- i indexes the down-sampling layer along the encoder
- j indexes the convolution layer of the dense block along the skip pathway
-
is a convolution operation followed by an activation function
-
denotes an up-sampling laye
-
denotes the concatenation laye
-
nodes at level j = 0 :一个输入,来自上一层encoder
-
nodes at level j = 1 : 两个输入,其中 1 个输入为同一 skip pathway 前1个节点的输出, 1 个为 lower skip pathway 的 up-sampled 输出
-
nodes at level j > 1 : j+1 个输入,其中 j 个输入为同一 skip pathway 前 j 个节点的输出, 1 个为 lower skip pathway 的 up-sampled 输出
Deep supervisio
enabling the model to operate in two mode:
- accurate mode
the outputs from all segmentation branches are averaged - fast mode
the final segmentation map is selected from only one of the segmentation branches , the choice of which determines the extent of model pruning and speed gain
Fig. 1c shows how the choice of segmentation branch in fast mode results in architectures of varying complexit