Loss for background class

HRNet / HRNet-Semantic-Segmentation

The OCR approach is rephrased as Segmentation Transformer: https://arxiv.org/abs/1909.11065. This is an official implementation of semantic segmentation for HRNet. https://arxiv.org/abs/1908.07919

Other

3.13k stars 686 forks source link

Loss for background class #267

Closed abhiagwl4262 closed 2 years ago

abhiagwl4262 commented 2 years ago

As much as I understood, IGNORE_LABEL=255 ignores the background class during loss computations. Am I correct?

If yes I think background shouldn't be ignored because it is important to differentiate between background and foreground.