search

fcdl94 / WILSON

Official implementation of "Incremental Learning in Semantic Segmentation from Image Labels"
https://arxiv.org/abs/2112.01882
MIT License
59 stars 9 forks source link

How to calculate MIoU #11

Closed Chongjie-Si closed 1 year ago

Chongjie-Si commented 1 year ago

Hi Fabio, I wonder how to compute the MIoU in your paper. Does 1-15 mean only calculating the average of 15 classes, and 16-20 mean the average of another 5 classes, and "all" mean that of total classes (including background)? Specifically, I want to know whether you included the IoU of the background when it comes to 16-20 and all.

fcdl94 commented 1 year ago

Hi! No, I did not included the background in 16-20 while it is included in the 'all'. Please, refer to Issue #2

Chongjie-Si commented 1 year ago

Thanks！