Bonus episode#

In the lecture, we mentioned N2V2. It can be used in 2D (not 3D yet...) by slightly changing the configuration, using the following parameters:

• blurpool=True
• skip_skipone=True
• n2v_manipulator="median"
• unet_residual=False

Try it out and compare with the previous result! Don't forget to save somewhere the data previously generated by the prediction notebook.