Table 5:
Overview of the used frameworks for the demonstrated tasks
Task | Dimension | Framework | Backbone |
---|---|---|---|
Label-free | 2D/3D | FCN, Pix2pix, CycleGAN | fnet, UNet, AttentionUnet, SwinUNETR, … |
Semantic segmentation | 2D/3D | FCN, CycleGAN | AttentionUnet, DynUnet, UNet3D |
Instance segmentation | 2D/3D | EmbedSeg | BranchedERFNet_2d, UNet3D |
Synthetic | 2D/3D | Pix2pix | AttentionUnet, fnet |
Denoising | 3D | FCN | UNet |
Modality transformation | 3D | FCN | UNet3D |
Staining transformation | 2D | Pix2pix | predefined_unet |