医学图像分割--U-Net: Convolutional Networks for Biomedical Image Segmentation

U-Net: Convolutional Networks for Biomedical Image Segmentation
https://arxiv.org/abs/1505.04597
Code: https://lmb.informatik.uni-freiburg.de/people/ronneber/u-net/

https://github.com/zhixuhao/unet
https://github.com/yihui-he/u-net
http://blog.leanote.com/post/zongwei/Pa
http://blog.csdn.net/hjimce/article/details/50268555

这里我们将 FCN 修改为 U-Net，主要是上采样阶段，我们同样也有许多特征通道，这样网络可以传递更多的 context 信息到 higher resolution 网络层
in the upsampling part we have also a large number of feature channels, which allow the network to propagate context information to higher resolution layers.

针对医学图像训练数据的问题，这里我们对训练数据使用 excessive data augmentation by applying elastic deformations
对于生物医学图像分割，elastic deformations 对于invariance的学习还是很有效的。

对于输入尺寸很大的图像，我们采用 Overlap-tile strategy

Ranking on the EM segmentation challenge

Result on the ISBI cell tracking challenge