Alice in AI Land

Hi, so how to get the result of the confusion matrix py~~,since I just get some metrics after the train_segmentor

train_segmentor(model, datasets, cfg, distributed=False, validate=True, meta=dict()) from tools.confusion_matrix import calculate_confusion_matrix calculate_confusion_matrix(datasets, results) but -------------------------------------------------------------- NameError ----> 1 calculate_confusion_matrix(datasets, results) NameError: name 'results' is not defined

Hi, could you show me the code of calculate precision~~, I got very large result when I add label = label[mask] matt=matthews_corrcoef(label, pred_label) matt=torch.tensor(matt) Matt.crack: 17278964.4800, Matt.non-crack: 964808.0000

Hello, what do you define the total_area_intersect, is it a TP+TN or TP? Thank you

Hi, as I save the best, How can I find the model that I saved~~ Thank you!!

It just said, Now best checkpoint is saved as best_mIoU_iter_200.pth. Where could I find the file,, Thank you~~

Oh, I find it , Thank you~

Hi, I have a folder contains test images for segmentation, I have train the model through train_segmentor(model, datasets, cfg, distributed=False, validate=True, meta=dict()) and save the best checkpoint is saved as...

Hi, after !python mmseg_custom/models/segmentors/__init__.py I get Traceback (most recent call last): File "mmseg_custom/models/segmentors/__init__.py", line 2, in from .encoder_decoder_mask2former import EncoderDecoderMask2Former ImportError: attempted relative import with no known parent package What...

when I type import mmseg_custom, there is ModuleNotFoundError: No module named 'mmdet' caused by ----> 3 from mmdet.utils import util_mixins