Table 2.

Number of trainable parameters in different parts of the UNet architectures.

Model	Encoder Arm	Bottleneck Layer	Decoder Arm	Intermediate Stages	FC	Classification Layer	Total Parameters
UNet	4685376	14157824	12188480	None	1154	6	31032840
UNet++	4685376	14157824	12593984	1275008	None	130	32712322
UNet3P	4685376	14157824	1196800	1447040	1154	6	21488200
Attention-UNet	4685376	14157824	12535680	523684	None	130	31902694

FC = fully connected layer.