Radial Basis Function Kernel for TPS Transformation (Wrong?)

[Atotem]

cp-vton-plus/networks.py

Line 291 in ff44e46

U = torch.mul(dist_squared, torch.log(dist_squared))

Hi!

I understand the Radial Basis kernel they are using is the following:
U = r^2 * log( r )

However, seems to me that the function used is this one:
U = r^2 * log( r^2 ), because after summing both pondered delta matrix, they don't apply root squared to the outcome.

Is there a reason to do it this way? I'm just trying to understand. I appreciate any insight.