Semantic Segmentation Models


Introduced by Ronneberger et al. in U-Net: Convolutional Networks for Biomedical Image Segmentation

U-Net is an architecture for semantic segmentation. It consists of a contracting path and an expansive path. The contracting path follows the typical architecture of a convolutional network. It consists of the repeated application of two 3x3 convolutions (unpadded convolutions), each followed by a rectified linear unit (ReLU) and a 2x2 max pooling operation with stride 2 for downsampling. At each downsampling step we double the number of feature channels. Every step in the expansive path consists of an upsampling of the feature map followed by a 2x2 convolution (“up-convolution”) that halves the number of feature channels, a concatenation with the correspondingly cropped feature map from the contracting path, and two 3x3 convolutions, each followed by a ReLU. The cropping is necessary due to the loss of border pixels in every convolution. At the final layer a 1x1 convolution is used to map each 64-component feature vector to the desired number of classes. In total the network has 23 convolutional layers.

Original MATLAB Code

Source: U-Net: Convolutional Networks for Biomedical Image Segmentation


Paper Code Results Date Stars


Task Papers Share
Semantic Segmentation 127 14.56%
Image Segmentation 98 11.24%
Medical Image Segmentation 62 7.11%
Denoising 40 4.59%
Tumor Segmentation 29 3.33%
Computed Tomography (CT) 21 2.41%
Image Generation 20 2.29%
Lesion Segmentation 19 2.18%
Super-Resolution 14 1.61%