I'm using this project to train my segmentation model. I find that the mask has a right-down offset to the image. Because the opencv resize_nearest is wrong. Please refer the opencv project issue:
https://github.com/opencv/opencv/issues/9096
https://github.com/opencv/opencv/issues/10146

The code of opencv is:
` for( x = 0; x < dsize.width; x++ )
{
int sx = cv

A follow up on SixLabors/ImageSharp#1378 (comment).

Currently 32 bit test execution is only done for .NET Framework, with dotnet xunit which is an obsolete tool today, we need to adapt dotnet test, and add 32 bit CI targets for both net5.0 and netcoreapp3.1. Opening an issue to remember and track this debt.

See:
https://twitter.com/tannergooding/s

See forum discussion

🐛 Bug

When computing the binary focal loss I got nan gradients on backward pass when torch.sig(prediction) equals 1. or 0. and 0.<gamma<1..

This probably does also concern the focal loss implementation.

To Reproduce

import torch
from kornia.losses import binary_focal_loss_with_logits

prediction=torch.tensor([[-100.,-100.,100.,100.]], requires_grad=True)