LappenX

LappenX t1_jbap16k wrote

That is exactly what should be happening. In the Cityscapes dataset for example you can always see part of the vehicle in the bottom of the image, and these pixels are set to ignore to be excluded from training loss and test metrics.

1