U-Net Encoder Block return variable name

Hello community,
I have a question regarding the U-Net encoder code.

In the U-Net: Encoder video, Laurence creates the endoer_block function, which returns both convolutional and pooling layers outputs.

The convolutional layer output variable is named f.
Why this name? Is it because of the “Feature map,” or is there another reason for the variable name election?

Thanks in advance!

Probably it is so, I din’t think much of it.

1 Like

I think Lawrence just chose for encoder block in relation to filter = f and pool_size = p. So I do not see any particular reason of choosing these variable name. Only another reason would be probably they wanted to specify which layer relates to filters and which to pool_size.

But don’t ask why there are not using d for dropout layer :joy: because dropout layer is usually related to pool_size layer only. Just kidding :wink:

This U-net assignment, Lawrence more tried to layer block names too, probably for a better presentation of assignment.



@gent.spah @Deepti_Prasad, thanks a lot for the answers.

I usually pay a lot of attention to the names I give to variables and functions, hence the question. I think properly crafted names can make a difference in maintainability.

Anyway, thanks again.

if you are a person with attention to detail, it is always a plus point.

1 Like