In the required format, what kind of prediction output should ”labeltrainids“ be?
In addition, What does ”labeltrainids_ invalid“ mean?
Hi,
Copying from the evaluation guidelines https://competitions.codalab.org/competitions/23553#learn_the_details-evaluation:
- labelTrainIds: predictions of semantic labels encoded using png images, where pixel values encode labels in Cityscapes trainIDs format according to Cityscapes documentation script https://github.com/mcordts/cityscapesScripts/blob/master/cityscapesscripts/helpers/labels.py. I.e., the labels should be uint8 integers from 0 to 18.
- labelTrainIds_invalid: predictions of semantic labels including the special label invalid, encoded using png images. Pixel values encode labels in Cityscapes trainIDs format (same as labelTrainIds) or the invalid label for the value 255. An invalid prediction for a pixel indicates that the model has not made a prediction for that pixel, typically due to low associated confidence. I.e., the labels should be uint8 integers from 0 to 18 or 255.
Posted by: sakaridis @ Jan. 12, 2022, 12:54 p.m.