Skip to main content
. 2022 Mar 12;12(3):698. doi: 10.3390/diagnostics12030698

Figure 2.

Figure 2

U-Net with two encoders where the features are combined using concatenation. The details of the Conv-block, MaxPool, and UpConv operations are similar to the U-Net with one encoder. For other feature fusion methods, namely add and weighted-sum, the concatenate operation was replaced by addition and weighted-sum, respectively.