A single separable-convolution 3D U-Net trained jointly on five segmentation datasets reaches mean Dice within about one point of per-task models while using around 1% of the parameters.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
eess.IV 1years
2019 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
3D U$^2$-Net: A 3D Universal U-Net for Multi-Domain Medical Image Segmentation
A single separable-convolution 3D U-Net trained jointly on five segmentation datasets reaches mean Dice within about one point of per-task models while using around 1% of the parameters.